]>
Commit | Line | Data |
---|---|---|
108b42b4 DH |
1 | ============================ |
2 | LINUX KERNEL MEMORY BARRIERS | |
3 | ============================ | |
4 | ||
5 | By: David Howells <dhowells@redhat.com> | |
90fddabf | 6 | Paul E. McKenney <paulmck@linux.vnet.ibm.com> |
e7720af5 PZ |
7 | Will Deacon <will.deacon@arm.com> |
8 | Peter Zijlstra <peterz@infradead.org> | |
108b42b4 | 9 | |
e7720af5 PZ |
10 | ========== |
11 | DISCLAIMER | |
12 | ========== | |
13 | ||
14 | This document is not a specification; it is intentionally (for the sake of | |
15 | brevity) and unintentionally (due to being human) incomplete. This document is | |
16 | meant as a guide to using the various memory barriers provided by Linux, but | |
621df431 AP |
17 | in case of any doubt (and there are many) please ask. Some doubts may be |
18 | resolved by referring to the formal memory consistency model and related | |
19 | documentation at tools/memory-model/. Nevertheless, even this memory | |
20 | model should be viewed as the collective opinion of its maintainers rather | |
21 | than as an infallible oracle. | |
e7720af5 PZ |
22 | |
23 | To repeat, this document is not a specification of what Linux expects from | |
24 | hardware. | |
25 | ||
8d4840e8 DH |
26 | The purpose of this document is twofold: |
27 | ||
28 | (1) to specify the minimum functionality that one can rely on for any | |
29 | particular barrier, and | |
30 | ||
31 | (2) to provide a guide as to how to use the barriers that are available. | |
32 | ||
33 | Note that an architecture can provide more than the minimum requirement | |
35bdc72a | 34 | for any particular barrier, but if the architecture provides less than |
8d4840e8 DH |
35 | that, that architecture is incorrect. |
36 | ||
37 | Note also that it is possible that a barrier may be a no-op for an | |
38 | architecture because the way that arch works renders an explicit barrier | |
39 | unnecessary in that case. | |
40 | ||
41 | ||
e7720af5 PZ |
42 | ======== |
43 | CONTENTS | |
44 | ======== | |
108b42b4 DH |
45 | |
46 | (*) Abstract memory access model. | |
47 | ||
48 | - Device operations. | |
49 | - Guarantees. | |
50 | ||
51 | (*) What are memory barriers? | |
52 | ||
53 | - Varieties of memory barrier. | |
54 | - What may not be assumed about memory barriers? | |
f28f0868 | 55 | - Data dependency barriers (historical). |
108b42b4 DH |
56 | - Control dependencies. |
57 | - SMP barrier pairing. | |
58 | - Examples of memory barrier sequences. | |
670bd95e | 59 | - Read memory barriers vs load speculation. |
f1ab25a3 | 60 | - Multicopy atomicity. |
108b42b4 DH |
61 | |
62 | (*) Explicit kernel barriers. | |
63 | ||
64 | - Compiler barrier. | |
81fc6323 | 65 | - CPU memory barriers. |
108b42b4 DH |
66 | - MMIO write barrier. |
67 | ||
68 | (*) Implicit kernel memory barriers. | |
69 | ||
166bda71 | 70 | - Lock acquisition functions. |
108b42b4 | 71 | - Interrupt disabling functions. |
50fa610a | 72 | - Sleep and wake-up functions. |
108b42b4 DH |
73 | - Miscellaneous functions. |
74 | ||
166bda71 | 75 | (*) Inter-CPU acquiring barrier effects. |
108b42b4 | 76 | |
166bda71 SP |
77 | - Acquires vs memory accesses. |
78 | - Acquires vs I/O accesses. | |
108b42b4 DH |
79 | |
80 | (*) Where are memory barriers needed? | |
81 | ||
82 | - Interprocessor interaction. | |
83 | - Atomic operations. | |
84 | - Accessing devices. | |
85 | - Interrupts. | |
86 | ||
87 | (*) Kernel I/O barrier effects. | |
88 | ||
89 | (*) Assumed minimum execution ordering model. | |
90 | ||
91 | (*) The effects of the cpu cache. | |
92 | ||
93 | - Cache coherency. | |
94 | - Cache coherency vs DMA. | |
95 | - Cache coherency vs MMIO. | |
96 | ||
97 | (*) The things CPUs get up to. | |
98 | ||
99 | - And then there's the Alpha. | |
01e1cd6d | 100 | - Virtual Machine Guests. |
108b42b4 | 101 | |
90fddabf DH |
102 | (*) Example uses. |
103 | ||
104 | - Circular buffers. | |
105 | ||
108b42b4 DH |
106 | (*) References. |
107 | ||
108 | ||
109 | ============================ | |
110 | ABSTRACT MEMORY ACCESS MODEL | |
111 | ============================ | |
112 | ||
113 | Consider the following abstract model of the system: | |
114 | ||
115 | : : | |
116 | : : | |
117 | : : | |
118 | +-------+ : +--------+ : +-------+ | |
119 | | | : | | : | | | |
120 | | | : | | : | | | |
121 | | CPU 1 |<----->| Memory |<----->| CPU 2 | | |
122 | | | : | | : | | | |
123 | | | : | | : | | | |
124 | +-------+ : +--------+ : +-------+ | |
125 | ^ : ^ : ^ | |
126 | | : | : | | |
127 | | : | : | | |
128 | | : v : | | |
129 | | : +--------+ : | | |
130 | | : | | : | | |
131 | | : | | : | | |
132 | +---------->| Device |<----------+ | |
133 | : | | : | |
134 | : | | : | |
135 | : +--------+ : | |
136 | : : | |
137 | ||
138 | Each CPU executes a program that generates memory access operations. In the | |
139 | abstract CPU, memory operation ordering is very relaxed, and a CPU may actually | |
140 | perform the memory operations in any order it likes, provided program causality | |
141 | appears to be maintained. Similarly, the compiler may also arrange the | |
142 | instructions it emits in any order it likes, provided it doesn't affect the | |
143 | apparent operation of the program. | |
144 | ||
145 | So in the above diagram, the effects of the memory operations performed by a | |
146 | CPU are perceived by the rest of the system as the operations cross the | |
147 | interface between the CPU and rest of the system (the dotted lines). | |
148 | ||
149 | ||
150 | For example, consider the following sequence of events: | |
151 | ||
152 | CPU 1 CPU 2 | |
153 | =============== =============== | |
154 | { A == 1; B == 2 } | |
615cc2c9 AD |
155 | A = 3; x = B; |
156 | B = 4; y = A; | |
108b42b4 DH |
157 | |
158 | The set of accesses as seen by the memory system in the middle can be arranged | |
159 | in 24 different combinations: | |
160 | ||
8ab8b3e1 PK |
161 | STORE A=3, STORE B=4, y=LOAD A->3, x=LOAD B->4 |
162 | STORE A=3, STORE B=4, x=LOAD B->4, y=LOAD A->3 | |
163 | STORE A=3, y=LOAD A->3, STORE B=4, x=LOAD B->4 | |
164 | STORE A=3, y=LOAD A->3, x=LOAD B->2, STORE B=4 | |
165 | STORE A=3, x=LOAD B->2, STORE B=4, y=LOAD A->3 | |
166 | STORE A=3, x=LOAD B->2, y=LOAD A->3, STORE B=4 | |
167 | STORE B=4, STORE A=3, y=LOAD A->3, x=LOAD B->4 | |
108b42b4 DH |
168 | STORE B=4, ... |
169 | ... | |
170 | ||
171 | and can thus result in four different combinations of values: | |
172 | ||
8ab8b3e1 PK |
173 | x == 2, y == 1 |
174 | x == 2, y == 3 | |
175 | x == 4, y == 1 | |
176 | x == 4, y == 3 | |
108b42b4 DH |
177 | |
178 | ||
179 | Furthermore, the stores committed by a CPU to the memory system may not be | |
180 | perceived by the loads made by another CPU in the same order as the stores were | |
181 | committed. | |
182 | ||
183 | ||
184 | As a further example, consider this sequence of events: | |
185 | ||
186 | CPU 1 CPU 2 | |
187 | =============== =============== | |
3dbf0913 | 188 | { A == 1, B == 2, C == 3, P == &A, Q == &C } |
108b42b4 DH |
189 | B = 4; Q = P; |
190 | P = &B D = *Q; | |
191 | ||
192 | There is an obvious data dependency here, as the value loaded into D depends on | |
193 | the address retrieved from P by CPU 2. At the end of the sequence, any of the | |
194 | following results are possible: | |
195 | ||
196 | (Q == &A) and (D == 1) | |
197 | (Q == &B) and (D == 2) | |
198 | (Q == &B) and (D == 4) | |
199 | ||
200 | Note that CPU 2 will never try and load C into D because the CPU will load P | |
201 | into Q before issuing the load of *Q. | |
202 | ||
203 | ||
204 | DEVICE OPERATIONS | |
205 | ----------------- | |
206 | ||
207 | Some devices present their control interfaces as collections of memory | |
208 | locations, but the order in which the control registers are accessed is very | |
209 | important. For instance, imagine an ethernet card with a set of internal | |
210 | registers that are accessed through an address port register (A) and a data | |
211 | port register (D). To read internal register 5, the following code might then | |
212 | be used: | |
213 | ||
214 | *A = 5; | |
215 | x = *D; | |
216 | ||
217 | but this might show up as either of the following two sequences: | |
218 | ||
219 | STORE *A = 5, x = LOAD *D | |
220 | x = LOAD *D, STORE *A = 5 | |
221 | ||
222 | the second of which will almost certainly result in a malfunction, since it set | |
223 | the address _after_ attempting to read the register. | |
224 | ||
225 | ||
226 | GUARANTEES | |
227 | ---------- | |
228 | ||
229 | There are some minimal guarantees that may be expected of a CPU: | |
230 | ||
231 | (*) On any given CPU, dependent memory accesses will be issued in order, with | |
232 | respect to itself. This means that for: | |
233 | ||
40555946 | 234 | Q = READ_ONCE(P); D = READ_ONCE(*Q); |
108b42b4 DH |
235 | |
236 | the CPU will issue the following memory operations: | |
237 | ||
238 | Q = LOAD P, D = LOAD *Q | |
239 | ||
40555946 PM |
240 | and always in that order. However, on DEC Alpha, READ_ONCE() also |
241 | emits a memory-barrier instruction, so that a DEC Alpha CPU will | |
242 | instead issue the following memory operations: | |
243 | ||
244 | Q = LOAD P, MEMORY_BARRIER, D = LOAD *Q, MEMORY_BARRIER | |
245 | ||
246 | Whether on DEC Alpha or not, the READ_ONCE() also prevents compiler | |
247 | mischief. | |
108b42b4 DH |
248 | |
249 | (*) Overlapping loads and stores within a particular CPU will appear to be | |
250 | ordered within that CPU. This means that for: | |
251 | ||
9af194ce | 252 | a = READ_ONCE(*X); WRITE_ONCE(*X, b); |
108b42b4 DH |
253 | |
254 | the CPU will only issue the following sequence of memory operations: | |
255 | ||
256 | a = LOAD *X, STORE *X = b | |
257 | ||
258 | And for: | |
259 | ||
9af194ce | 260 | WRITE_ONCE(*X, c); d = READ_ONCE(*X); |
108b42b4 DH |
261 | |
262 | the CPU will only issue: | |
263 | ||
264 | STORE *X = c, d = LOAD *X | |
265 | ||
fa00e7e1 | 266 | (Loads and stores overlap if they are targeted at overlapping pieces of |
108b42b4 DH |
267 | memory). |
268 | ||
269 | And there are a number of things that _must_ or _must_not_ be assumed: | |
270 | ||
9af194ce PM |
271 | (*) It _must_not_ be assumed that the compiler will do what you want |
272 | with memory references that are not protected by READ_ONCE() and | |
273 | WRITE_ONCE(). Without them, the compiler is within its rights to | |
274 | do all sorts of "creative" transformations, which are covered in | |
895f5542 | 275 | the COMPILER BARRIER section. |
2ecf8101 | 276 | |
108b42b4 DH |
277 | (*) It _must_not_ be assumed that independent loads and stores will be issued |
278 | in the order given. This means that for: | |
279 | ||
280 | X = *A; Y = *B; *D = Z; | |
281 | ||
282 | we may get any of the following sequences: | |
283 | ||
284 | X = LOAD *A, Y = LOAD *B, STORE *D = Z | |
285 | X = LOAD *A, STORE *D = Z, Y = LOAD *B | |
286 | Y = LOAD *B, X = LOAD *A, STORE *D = Z | |
287 | Y = LOAD *B, STORE *D = Z, X = LOAD *A | |
288 | STORE *D = Z, X = LOAD *A, Y = LOAD *B | |
289 | STORE *D = Z, Y = LOAD *B, X = LOAD *A | |
290 | ||
291 | (*) It _must_ be assumed that overlapping memory accesses may be merged or | |
292 | discarded. This means that for: | |
293 | ||
294 | X = *A; Y = *(A + 4); | |
295 | ||
296 | we may get any one of the following sequences: | |
297 | ||
298 | X = LOAD *A; Y = LOAD *(A + 4); | |
299 | Y = LOAD *(A + 4); X = LOAD *A; | |
300 | {X, Y} = LOAD {*A, *(A + 4) }; | |
301 | ||
302 | And for: | |
303 | ||
f191eec5 | 304 | *A = X; *(A + 4) = Y; |
108b42b4 | 305 | |
f191eec5 | 306 | we may get any of: |
108b42b4 | 307 | |
f191eec5 PM |
308 | STORE *A = X; STORE *(A + 4) = Y; |
309 | STORE *(A + 4) = Y; STORE *A = X; | |
310 | STORE {*A, *(A + 4) } = {X, Y}; | |
108b42b4 | 311 | |
432fbf3c PM |
312 | And there are anti-guarantees: |
313 | ||
314 | (*) These guarantees do not apply to bitfields, because compilers often | |
315 | generate code to modify these using non-atomic read-modify-write | |
316 | sequences. Do not attempt to use bitfields to synchronize parallel | |
317 | algorithms. | |
318 | ||
319 | (*) Even in cases where bitfields are protected by locks, all fields | |
320 | in a given bitfield must be protected by one lock. If two fields | |
321 | in a given bitfield are protected by different locks, the compiler's | |
322 | non-atomic read-modify-write sequences can cause an update to one | |
323 | field to corrupt the value of an adjacent field. | |
324 | ||
325 | (*) These guarantees apply only to properly aligned and sized scalar | |
326 | variables. "Properly sized" currently means variables that are | |
327 | the same size as "char", "short", "int" and "long". "Properly | |
328 | aligned" means the natural alignment, thus no constraints for | |
329 | "char", two-byte alignment for "short", four-byte alignment for | |
330 | "int", and either four-byte or eight-byte alignment for "long", | |
331 | on 32-bit and 64-bit systems, respectively. Note that these | |
332 | guarantees were introduced into the C11 standard, so beware when | |
333 | using older pre-C11 compilers (for example, gcc 4.6). The portion | |
334 | of the standard containing this guarantee is Section 3.14, which | |
335 | defines "memory location" as follows: | |
336 | ||
337 | memory location | |
338 | either an object of scalar type, or a maximal sequence | |
339 | of adjacent bit-fields all having nonzero width | |
340 | ||
341 | NOTE 1: Two threads of execution can update and access | |
342 | separate memory locations without interfering with | |
343 | each other. | |
344 | ||
345 | NOTE 2: A bit-field and an adjacent non-bit-field member | |
346 | are in separate memory locations. The same applies | |
347 | to two bit-fields, if one is declared inside a nested | |
348 | structure declaration and the other is not, or if the two | |
349 | are separated by a zero-length bit-field declaration, | |
350 | or if they are separated by a non-bit-field member | |
351 | declaration. It is not safe to concurrently update two | |
352 | bit-fields in the same structure if all members declared | |
353 | between them are also bit-fields, no matter what the | |
354 | sizes of those intervening bit-fields happen to be. | |
355 | ||
108b42b4 DH |
356 | |
357 | ========================= | |
358 | WHAT ARE MEMORY BARRIERS? | |
359 | ========================= | |
360 | ||
361 | As can be seen above, independent memory operations are effectively performed | |
362 | in random order, but this can be a problem for CPU-CPU interaction and for I/O. | |
363 | What is required is some way of intervening to instruct the compiler and the | |
364 | CPU to restrict the order. | |
365 | ||
366 | Memory barriers are such interventions. They impose a perceived partial | |
2b94895b DH |
367 | ordering over the memory operations on either side of the barrier. |
368 | ||
369 | Such enforcement is important because the CPUs and other devices in a system | |
81fc6323 | 370 | can use a variety of tricks to improve performance, including reordering, |
2b94895b DH |
371 | deferral and combination of memory operations; speculative loads; speculative |
372 | branch prediction and various types of caching. Memory barriers are used to | |
373 | override or suppress these tricks, allowing the code to sanely control the | |
374 | interaction of multiple CPUs and/or devices. | |
108b42b4 DH |
375 | |
376 | ||
377 | VARIETIES OF MEMORY BARRIER | |
378 | --------------------------- | |
379 | ||
380 | Memory barriers come in four basic varieties: | |
381 | ||
382 | (1) Write (or store) memory barriers. | |
383 | ||
384 | A write memory barrier gives a guarantee that all the STORE operations | |
385 | specified before the barrier will appear to happen before all the STORE | |
386 | operations specified after the barrier with respect to the other | |
387 | components of the system. | |
388 | ||
389 | A write barrier is a partial ordering on stores only; it is not required | |
390 | to have any effect on loads. | |
391 | ||
6bc39274 | 392 | A CPU can be viewed as committing a sequence of store operations to the |
5692fcc6 GP |
393 | memory system as time progresses. All stores _before_ a write barrier |
394 | will occur _before_ all the stores after the write barrier. | |
108b42b4 DH |
395 | |
396 | [!] Note that write barriers should normally be paired with read or data | |
397 | dependency barriers; see the "SMP barrier pairing" subsection. | |
398 | ||
399 | ||
400 | (2) Data dependency barriers. | |
401 | ||
402 | A data dependency barrier is a weaker form of read barrier. In the case | |
403 | where two loads are performed such that the second depends on the result | |
404 | of the first (eg: the first load retrieves the address to which the second | |
405 | load will be directed), a data dependency barrier would be required to | |
51de7889 | 406 | make sure that the target of the second load is updated after the address |
108b42b4 DH |
407 | obtained by the first load is accessed. |
408 | ||
409 | A data dependency barrier is a partial ordering on interdependent loads | |
410 | only; it is not required to have any effect on stores, independent loads | |
411 | or overlapping loads. | |
412 | ||
413 | As mentioned in (1), the other CPUs in the system can be viewed as | |
414 | committing sequences of stores to the memory system that the CPU being | |
415 | considered can then perceive. A data dependency barrier issued by the CPU | |
416 | under consideration guarantees that for any load preceding it, if that | |
417 | load touches one of a sequence of stores from another CPU, then by the | |
418 | time the barrier completes, the effects of all the stores prior to that | |
419 | touched by the load will be perceptible to any loads issued after the data | |
420 | dependency barrier. | |
421 | ||
422 | See the "Examples of memory barrier sequences" subsection for diagrams | |
423 | showing the ordering constraints. | |
424 | ||
425 | [!] Note that the first load really has to have a _data_ dependency and | |
426 | not a control dependency. If the address for the second load is dependent | |
427 | on the first load, but the dependency is through a conditional rather than | |
428 | actually loading the address itself, then it's a _control_ dependency and | |
429 | a full read barrier or better is required. See the "Control dependencies" | |
430 | subsection for more information. | |
431 | ||
432 | [!] Note that data dependency barriers should normally be paired with | |
433 | write barriers; see the "SMP barrier pairing" subsection. | |
434 | ||
435 | ||
436 | (3) Read (or load) memory barriers. | |
437 | ||
438 | A read barrier is a data dependency barrier plus a guarantee that all the | |
439 | LOAD operations specified before the barrier will appear to happen before | |
440 | all the LOAD operations specified after the barrier with respect to the | |
441 | other components of the system. | |
442 | ||
443 | A read barrier is a partial ordering on loads only; it is not required to | |
444 | have any effect on stores. | |
445 | ||
446 | Read memory barriers imply data dependency barriers, and so can substitute | |
447 | for them. | |
448 | ||
449 | [!] Note that read barriers should normally be paired with write barriers; | |
450 | see the "SMP barrier pairing" subsection. | |
451 | ||
452 | ||
453 | (4) General memory barriers. | |
454 | ||
670bd95e DH |
455 | A general memory barrier gives a guarantee that all the LOAD and STORE |
456 | operations specified before the barrier will appear to happen before all | |
457 | the LOAD and STORE operations specified after the barrier with respect to | |
458 | the other components of the system. | |
459 | ||
460 | A general memory barrier is a partial ordering over both loads and stores. | |
108b42b4 DH |
461 | |
462 | General memory barriers imply both read and write memory barriers, and so | |
463 | can substitute for either. | |
464 | ||
465 | ||
466 | And a couple of implicit varieties: | |
467 | ||
2e4f5382 | 468 | (5) ACQUIRE operations. |
108b42b4 DH |
469 | |
470 | This acts as a one-way permeable barrier. It guarantees that all memory | |
2e4f5382 PZ |
471 | operations after the ACQUIRE operation will appear to happen after the |
472 | ACQUIRE operation with respect to the other components of the system. | |
787df638 | 473 | ACQUIRE operations include LOCK operations and both smp_load_acquire() |
2f359c7e | 474 | and smp_cond_load_acquire() operations. |
108b42b4 | 475 | |
2e4f5382 PZ |
476 | Memory operations that occur before an ACQUIRE operation may appear to |
477 | happen after it completes. | |
108b42b4 | 478 | |
2e4f5382 PZ |
479 | An ACQUIRE operation should almost always be paired with a RELEASE |
480 | operation. | |
108b42b4 DH |
481 | |
482 | ||
2e4f5382 | 483 | (6) RELEASE operations. |
108b42b4 DH |
484 | |
485 | This also acts as a one-way permeable barrier. It guarantees that all | |
2e4f5382 PZ |
486 | memory operations before the RELEASE operation will appear to happen |
487 | before the RELEASE operation with respect to the other components of the | |
488 | system. RELEASE operations include UNLOCK operations and | |
489 | smp_store_release() operations. | |
108b42b4 | 490 | |
2e4f5382 | 491 | Memory operations that occur after a RELEASE operation may appear to |
108b42b4 DH |
492 | happen before it completes. |
493 | ||
2e4f5382 PZ |
494 | The use of ACQUIRE and RELEASE operations generally precludes the need |
495 | for other sorts of memory barrier (but note the exceptions mentioned in | |
496 | the subsection "MMIO write barrier"). In addition, a RELEASE+ACQUIRE | |
497 | pair is -not- guaranteed to act as a full memory barrier. However, after | |
498 | an ACQUIRE on a given variable, all memory accesses preceding any prior | |
499 | RELEASE on that same variable are guaranteed to be visible. In other | |
500 | words, within a given variable's critical section, all accesses of all | |
501 | previous critical sections for that variable are guaranteed to have | |
502 | completed. | |
17eb88e0 | 503 | |
2e4f5382 PZ |
504 | This means that ACQUIRE acts as a minimal "acquire" operation and |
505 | RELEASE acts as a minimal "release" operation. | |
108b42b4 | 506 | |
706eeb3e PZ |
507 | A subset of the atomic operations described in atomic_t.txt have ACQUIRE and |
508 | RELEASE variants in addition to fully-ordered and relaxed (no barrier | |
509 | semantics) definitions. For compound atomics performing both a load and a | |
510 | store, ACQUIRE semantics apply only to the load and RELEASE semantics apply | |
511 | only to the store portion of the operation. | |
108b42b4 DH |
512 | |
513 | Memory barriers are only required where there's a possibility of interaction | |
514 | between two CPUs or between a CPU and a device. If it can be guaranteed that | |
515 | there won't be any such interaction in any particular piece of code, then | |
516 | memory barriers are unnecessary in that piece of code. | |
517 | ||
518 | ||
519 | Note that these are the _minimum_ guarantees. Different architectures may give | |
520 | more substantial guarantees, but they may _not_ be relied upon outside of arch | |
521 | specific code. | |
522 | ||
523 | ||
524 | WHAT MAY NOT BE ASSUMED ABOUT MEMORY BARRIERS? | |
525 | ---------------------------------------------- | |
526 | ||
527 | There are certain things that the Linux kernel memory barriers do not guarantee: | |
528 | ||
529 | (*) There is no guarantee that any of the memory accesses specified before a | |
530 | memory barrier will be _complete_ by the completion of a memory barrier | |
531 | instruction; the barrier can be considered to draw a line in that CPU's | |
532 | access queue that accesses of the appropriate type may not cross. | |
533 | ||
534 | (*) There is no guarantee that issuing a memory barrier on one CPU will have | |
535 | any direct effect on another CPU or any other hardware in the system. The | |
536 | indirect effect will be the order in which the second CPU sees the effects | |
537 | of the first CPU's accesses occur, but see the next point: | |
538 | ||
6bc39274 | 539 | (*) There is no guarantee that a CPU will see the correct order of effects |
108b42b4 DH |
540 | from a second CPU's accesses, even _if_ the second CPU uses a memory |
541 | barrier, unless the first CPU _also_ uses a matching memory barrier (see | |
542 | the subsection on "SMP Barrier Pairing"). | |
543 | ||
544 | (*) There is no guarantee that some intervening piece of off-the-CPU | |
545 | hardware[*] will not reorder the memory accesses. CPU cache coherency | |
546 | mechanisms should propagate the indirect effects of a memory barrier | |
547 | between CPUs, but might not do so in order. | |
548 | ||
549 | [*] For information on bus mastering DMA and coherency please read: | |
550 | ||
4b5ff469 | 551 | Documentation/PCI/pci.txt |
395cf969 | 552 | Documentation/DMA-API-HOWTO.txt |
108b42b4 DH |
553 | Documentation/DMA-API.txt |
554 | ||
555 | ||
f28f0868 PM |
556 | DATA DEPENDENCY BARRIERS (HISTORICAL) |
557 | ------------------------------------- | |
558 | ||
559 | As of v4.15 of the Linux kernel, an smp_read_barrier_depends() was | |
560 | added to READ_ONCE(), which means that about the only people who | |
561 | need to pay attention to this section are those working on DEC Alpha | |
562 | architecture-specific code and those working on READ_ONCE() itself. | |
563 | For those who need it, and for those who are interested in the history, | |
564 | here is the story of data-dependency barriers. | |
108b42b4 DH |
565 | |
566 | The usage requirements of data dependency barriers are a little subtle, and | |
567 | it's not always obvious that they're needed. To illustrate, consider the | |
568 | following sequence of events: | |
569 | ||
2ecf8101 PM |
570 | CPU 1 CPU 2 |
571 | =============== =============== | |
3dbf0913 | 572 | { A == 1, B == 2, C == 3, P == &A, Q == &C } |
108b42b4 DH |
573 | B = 4; |
574 | <write barrier> | |
9af194ce PM |
575 | WRITE_ONCE(P, &B) |
576 | Q = READ_ONCE(P); | |
2ecf8101 | 577 | D = *Q; |
108b42b4 DH |
578 | |
579 | There's a clear data dependency here, and it would seem that by the end of the | |
580 | sequence, Q must be either &A or &B, and that: | |
581 | ||
582 | (Q == &A) implies (D == 1) | |
583 | (Q == &B) implies (D == 4) | |
584 | ||
81fc6323 | 585 | But! CPU 2's perception of P may be updated _before_ its perception of B, thus |
108b42b4 DH |
586 | leading to the following situation: |
587 | ||
588 | (Q == &B) and (D == 2) ???? | |
589 | ||
590 | Whilst this may seem like a failure of coherency or causality maintenance, it | |
591 | isn't, and this behaviour can be observed on certain real CPUs (such as the DEC | |
592 | Alpha). | |
593 | ||
2b94895b DH |
594 | To deal with this, a data dependency barrier or better must be inserted |
595 | between the address load and the data load: | |
108b42b4 | 596 | |
2ecf8101 PM |
597 | CPU 1 CPU 2 |
598 | =============== =============== | |
3dbf0913 | 599 | { A == 1, B == 2, C == 3, P == &A, Q == &C } |
108b42b4 DH |
600 | B = 4; |
601 | <write barrier> | |
9af194ce PM |
602 | WRITE_ONCE(P, &B); |
603 | Q = READ_ONCE(P); | |
2ecf8101 PM |
604 | <data dependency barrier> |
605 | D = *Q; | |
108b42b4 DH |
606 | |
607 | This enforces the occurrence of one of the two implications, and prevents the | |
608 | third possibility from arising. | |
609 | ||
66ce3a4d PM |
610 | |
611 | [!] Note that this extremely counterintuitive situation arises most easily on | |
612 | machines with split caches, so that, for example, one cache bank processes | |
613 | even-numbered cache lines and the other bank processes odd-numbered cache | |
614 | lines. The pointer P might be stored in an odd-numbered cache line, and the | |
615 | variable B might be stored in an even-numbered cache line. Then, if the | |
616 | even-numbered bank of the reading CPU's cache is extremely busy while the | |
617 | odd-numbered bank is idle, one can see the new value of the pointer P (&B), | |
618 | but the old value of the variable B (2). | |
619 | ||
620 | ||
621 | A data-dependency barrier is not required to order dependent writes | |
622 | because the CPUs that the Linux kernel supports don't do writes | |
623 | until they are certain (1) that the write will actually happen, (2) | |
624 | of the location of the write, and (3) of the value to be written. | |
625 | But please carefully read the "CONTROL DEPENDENCIES" section and the | |
626 | Documentation/RCU/rcu_dereference.txt file: The compiler can and does | |
627 | break dependencies in a great many highly creative ways. | |
92a84dd2 PM |
628 | |
629 | CPU 1 CPU 2 | |
630 | =============== =============== | |
631 | { A == 1, B == 2, C = 3, P == &A, Q == &C } | |
632 | B = 4; | |
633 | <write barrier> | |
634 | WRITE_ONCE(P, &B); | |
635 | Q = READ_ONCE(P); | |
66ce3a4d | 636 | WRITE_ONCE(*Q, 5); |
92a84dd2 | 637 | |
66ce3a4d PM |
638 | Therefore, no data-dependency barrier is required to order the read into |
639 | Q with the store into *Q. In other words, this outcome is prohibited, | |
640 | even without a data-dependency barrier: | |
92a84dd2 | 641 | |
8b9e7715 | 642 | (Q == &B) && (B == 4) |
92a84dd2 PM |
643 | |
644 | Please note that this pattern should be rare. After all, the whole point | |
645 | of dependency ordering is to -prevent- writes to the data structure, along | |
646 | with the expensive cache misses associated with those writes. This pattern | |
66ce3a4d PM |
647 | can be used to record rare error conditions and the like, and the CPUs' |
648 | naturally occurring ordering prevents such records from being lost. | |
108b42b4 DH |
649 | |
650 | ||
f1ab25a3 PM |
651 | Note well that the ordering provided by a data dependency is local to |
652 | the CPU containing it. See the section on "Multicopy atomicity" for | |
653 | more information. | |
654 | ||
655 | ||
2ecf8101 PM |
656 | The data dependency barrier is very important to the RCU system, |
657 | for example. See rcu_assign_pointer() and rcu_dereference() in | |
658 | include/linux/rcupdate.h. This permits the current target of an RCU'd | |
659 | pointer to be replaced with a new modified target, without the replacement | |
660 | target appearing to be incompletely initialised. | |
108b42b4 DH |
661 | |
662 | See also the subsection on "Cache Coherency" for a more thorough example. | |
663 | ||
664 | ||
665 | CONTROL DEPENDENCIES | |
666 | -------------------- | |
667 | ||
c8241f85 PM |
668 | Control dependencies can be a bit tricky because current compilers do |
669 | not understand them. The purpose of this section is to help you prevent | |
670 | the compiler's ignorance from breaking your code. | |
671 | ||
ff382810 PM |
672 | A load-load control dependency requires a full read memory barrier, not |
673 | simply a data dependency barrier to make it work correctly. Consider the | |
674 | following bit of code: | |
108b42b4 | 675 | |
9af194ce | 676 | q = READ_ONCE(a); |
18c03c61 PZ |
677 | if (q) { |
678 | <data dependency barrier> /* BUG: No data dependency!!! */ | |
9af194ce | 679 | p = READ_ONCE(b); |
45c8a36a | 680 | } |
108b42b4 DH |
681 | |
682 | This will not have the desired effect because there is no actual data | |
2ecf8101 PM |
683 | dependency, but rather a control dependency that the CPU may short-circuit |
684 | by attempting to predict the outcome in advance, so that other CPUs see | |
685 | the load from b as having happened before the load from a. In such a | |
686 | case what's actually required is: | |
108b42b4 | 687 | |
9af194ce | 688 | q = READ_ONCE(a); |
18c03c61 | 689 | if (q) { |
45c8a36a | 690 | <read barrier> |
9af194ce | 691 | p = READ_ONCE(b); |
45c8a36a | 692 | } |
18c03c61 PZ |
693 | |
694 | However, stores are not speculated. This means that ordering -is- provided | |
ff382810 | 695 | for load-store control dependencies, as in the following example: |
18c03c61 | 696 | |
105ff3cb | 697 | q = READ_ONCE(a); |
18c03c61 | 698 | if (q) { |
c8241f85 | 699 | WRITE_ONCE(b, 1); |
18c03c61 PZ |
700 | } |
701 | ||
c8241f85 PM |
702 | Control dependencies pair normally with other types of barriers. |
703 | That said, please note that neither READ_ONCE() nor WRITE_ONCE() | |
704 | are optional! Without the READ_ONCE(), the compiler might combine the | |
705 | load from 'a' with other loads from 'a'. Without the WRITE_ONCE(), | |
706 | the compiler might combine the store to 'b' with other stores to 'b'. | |
707 | Either can result in highly counterintuitive effects on ordering. | |
18c03c61 PZ |
708 | |
709 | Worse yet, if the compiler is able to prove (say) that the value of | |
710 | variable 'a' is always non-zero, it would be well within its rights | |
711 | to optimize the original example by eliminating the "if" statement | |
712 | as follows: | |
713 | ||
714 | q = a; | |
c8241f85 | 715 | b = 1; /* BUG: Compiler and CPU can both reorder!!! */ |
2456d2a6 | 716 | |
105ff3cb | 717 | So don't leave out the READ_ONCE(). |
18c03c61 | 718 | |
2456d2a6 PM |
719 | It is tempting to try to enforce ordering on identical stores on both |
720 | branches of the "if" statement as follows: | |
18c03c61 | 721 | |
105ff3cb | 722 | q = READ_ONCE(a); |
18c03c61 | 723 | if (q) { |
9b2b3bf5 | 724 | barrier(); |
c8241f85 | 725 | WRITE_ONCE(b, 1); |
18c03c61 PZ |
726 | do_something(); |
727 | } else { | |
9b2b3bf5 | 728 | barrier(); |
c8241f85 | 729 | WRITE_ONCE(b, 1); |
18c03c61 PZ |
730 | do_something_else(); |
731 | } | |
732 | ||
2456d2a6 PM |
733 | Unfortunately, current compilers will transform this as follows at high |
734 | optimization levels: | |
18c03c61 | 735 | |
105ff3cb | 736 | q = READ_ONCE(a); |
2456d2a6 | 737 | barrier(); |
c8241f85 | 738 | WRITE_ONCE(b, 1); /* BUG: No ordering vs. load from a!!! */ |
18c03c61 | 739 | if (q) { |
c8241f85 | 740 | /* WRITE_ONCE(b, 1); -- moved up, BUG!!! */ |
18c03c61 PZ |
741 | do_something(); |
742 | } else { | |
c8241f85 | 743 | /* WRITE_ONCE(b, 1); -- moved up, BUG!!! */ |
18c03c61 PZ |
744 | do_something_else(); |
745 | } | |
746 | ||
2456d2a6 PM |
747 | Now there is no conditional between the load from 'a' and the store to |
748 | 'b', which means that the CPU is within its rights to reorder them: | |
749 | The conditional is absolutely required, and must be present in the | |
750 | assembly code even after all compiler optimizations have been applied. | |
751 | Therefore, if you need ordering in this example, you need explicit | |
752 | memory barriers, for example, smp_store_release(): | |
18c03c61 | 753 | |
9af194ce | 754 | q = READ_ONCE(a); |
2456d2a6 | 755 | if (q) { |
c8241f85 | 756 | smp_store_release(&b, 1); |
18c03c61 PZ |
757 | do_something(); |
758 | } else { | |
c8241f85 | 759 | smp_store_release(&b, 1); |
18c03c61 PZ |
760 | do_something_else(); |
761 | } | |
762 | ||
2456d2a6 PM |
763 | In contrast, without explicit memory barriers, two-legged-if control |
764 | ordering is guaranteed only when the stores differ, for example: | |
765 | ||
105ff3cb | 766 | q = READ_ONCE(a); |
2456d2a6 | 767 | if (q) { |
c8241f85 | 768 | WRITE_ONCE(b, 1); |
2456d2a6 PM |
769 | do_something(); |
770 | } else { | |
c8241f85 | 771 | WRITE_ONCE(b, 2); |
2456d2a6 PM |
772 | do_something_else(); |
773 | } | |
774 | ||
105ff3cb LT |
775 | The initial READ_ONCE() is still required to prevent the compiler from |
776 | proving the value of 'a'. | |
18c03c61 PZ |
777 | |
778 | In addition, you need to be careful what you do with the local variable 'q', | |
779 | otherwise the compiler might be able to guess the value and again remove | |
780 | the needed conditional. For example: | |
781 | ||
105ff3cb | 782 | q = READ_ONCE(a); |
18c03c61 | 783 | if (q % MAX) { |
c8241f85 | 784 | WRITE_ONCE(b, 1); |
18c03c61 PZ |
785 | do_something(); |
786 | } else { | |
c8241f85 | 787 | WRITE_ONCE(b, 2); |
18c03c61 PZ |
788 | do_something_else(); |
789 | } | |
790 | ||
791 | If MAX is defined to be 1, then the compiler knows that (q % MAX) is | |
792 | equal to zero, in which case the compiler is within its rights to | |
793 | transform the above code into the following: | |
794 | ||
105ff3cb | 795 | q = READ_ONCE(a); |
b26cfc48 | 796 | WRITE_ONCE(b, 2); |
18c03c61 PZ |
797 | do_something_else(); |
798 | ||
2456d2a6 PM |
799 | Given this transformation, the CPU is not required to respect the ordering |
800 | between the load from variable 'a' and the store to variable 'b'. It is | |
801 | tempting to add a barrier(), but this does not help. The conditional | |
802 | is gone, and the barrier won't bring it back. Therefore, if you are | |
803 | relying on this ordering, you should make sure that MAX is greater than | |
804 | one, perhaps as follows: | |
18c03c61 | 805 | |
105ff3cb | 806 | q = READ_ONCE(a); |
18c03c61 PZ |
807 | BUILD_BUG_ON(MAX <= 1); /* Order load from a with store to b. */ |
808 | if (q % MAX) { | |
c8241f85 | 809 | WRITE_ONCE(b, 1); |
18c03c61 PZ |
810 | do_something(); |
811 | } else { | |
c8241f85 | 812 | WRITE_ONCE(b, 2); |
18c03c61 PZ |
813 | do_something_else(); |
814 | } | |
815 | ||
2456d2a6 PM |
816 | Please note once again that the stores to 'b' differ. If they were |
817 | identical, as noted earlier, the compiler could pull this store outside | |
818 | of the 'if' statement. | |
819 | ||
8b19d1de PM |
820 | You must also be careful not to rely too much on boolean short-circuit |
821 | evaluation. Consider this example: | |
822 | ||
105ff3cb | 823 | q = READ_ONCE(a); |
57aecae9 | 824 | if (q || 1 > 0) |
9af194ce | 825 | WRITE_ONCE(b, 1); |
8b19d1de | 826 | |
5af4692a PM |
827 | Because the first condition cannot fault and the second condition is |
828 | always true, the compiler can transform this example as following, | |
829 | defeating control dependency: | |
8b19d1de | 830 | |
105ff3cb | 831 | q = READ_ONCE(a); |
9af194ce | 832 | WRITE_ONCE(b, 1); |
8b19d1de PM |
833 | |
834 | This example underscores the need to ensure that the compiler cannot | |
9af194ce | 835 | out-guess your code. More generally, although READ_ONCE() does force |
8b19d1de PM |
836 | the compiler to actually emit code for a given load, it does not force |
837 | the compiler to use the results. | |
838 | ||
ebff09a6 PM |
839 | In addition, control dependencies apply only to the then-clause and |
840 | else-clause of the if-statement in question. In particular, it does | |
841 | not necessarily apply to code following the if-statement: | |
842 | ||
843 | q = READ_ONCE(a); | |
844 | if (q) { | |
c8241f85 | 845 | WRITE_ONCE(b, 1); |
ebff09a6 | 846 | } else { |
c8241f85 | 847 | WRITE_ONCE(b, 2); |
ebff09a6 | 848 | } |
c8241f85 | 849 | WRITE_ONCE(c, 1); /* BUG: No ordering against the read from 'a'. */ |
ebff09a6 PM |
850 | |
851 | It is tempting to argue that there in fact is ordering because the | |
852 | compiler cannot reorder volatile accesses and also cannot reorder | |
c8241f85 PM |
853 | the writes to 'b' with the condition. Unfortunately for this line |
854 | of reasoning, the compiler might compile the two writes to 'b' as | |
ebff09a6 PM |
855 | conditional-move instructions, as in this fanciful pseudo-assembly |
856 | language: | |
857 | ||
858 | ld r1,a | |
ebff09a6 | 859 | cmp r1,$0 |
c8241f85 PM |
860 | cmov,ne r4,$1 |
861 | cmov,eq r4,$2 | |
ebff09a6 PM |
862 | st r4,b |
863 | st $1,c | |
864 | ||
865 | A weakly ordered CPU would have no dependency of any sort between the load | |
c8241f85 | 866 | from 'a' and the store to 'c'. The control dependencies would extend |
ebff09a6 PM |
867 | only to the pair of cmov instructions and the store depending on them. |
868 | In short, control dependencies apply only to the stores in the then-clause | |
869 | and else-clause of the if-statement in question (including functions | |
870 | invoked by those two clauses), not to code following that if-statement. | |
871 | ||
18c03c61 | 872 | |
f1ab25a3 PM |
873 | Note well that the ordering provided by a control dependency is local |
874 | to the CPU containing it. See the section on "Multicopy atomicity" | |
875 | for more information. | |
18c03c61 | 876 | |
18c03c61 PZ |
877 | |
878 | In summary: | |
879 | ||
880 | (*) Control dependencies can order prior loads against later stores. | |
881 | However, they do -not- guarantee any other sort of ordering: | |
882 | Not prior loads against later loads, nor prior stores against | |
883 | later anything. If you need these other forms of ordering, | |
d87510c5 | 884 | use smp_rmb(), smp_wmb(), or, in the case of prior stores and |
18c03c61 PZ |
885 | later loads, smp_mb(). |
886 | ||
7817b799 PM |
887 | (*) If both legs of the "if" statement begin with identical stores to |
888 | the same variable, then those stores must be ordered, either by | |
889 | preceding both of them with smp_mb() or by using smp_store_release() | |
890 | to carry out the stores. Please note that it is -not- sufficient | |
a5052657 PM |
891 | to use barrier() at beginning of each leg of the "if" statement |
892 | because, as shown by the example above, optimizing compilers can | |
893 | destroy the control dependency while respecting the letter of the | |
894 | barrier() law. | |
9b2b3bf5 | 895 | |
18c03c61 | 896 | (*) Control dependencies require at least one run-time conditional |
586dd56a | 897 | between the prior load and the subsequent store, and this |
9af194ce PM |
898 | conditional must involve the prior load. If the compiler is able |
899 | to optimize the conditional away, it will have also optimized | |
105ff3cb LT |
900 | away the ordering. Careful use of READ_ONCE() and WRITE_ONCE() |
901 | can help to preserve the needed conditional. | |
18c03c61 PZ |
902 | |
903 | (*) Control dependencies require that the compiler avoid reordering the | |
105ff3cb LT |
904 | dependency into nonexistence. Careful use of READ_ONCE() or |
905 | atomic{,64}_read() can help to preserve your control dependency. | |
895f5542 | 906 | Please see the COMPILER BARRIER section for more information. |
18c03c61 | 907 | |
ebff09a6 PM |
908 | (*) Control dependencies apply only to the then-clause and else-clause |
909 | of the if-statement containing the control dependency, including | |
910 | any functions that these two clauses call. Control dependencies | |
911 | do -not- apply to code following the if-statement containing the | |
912 | control dependency. | |
913 | ||
ff382810 PM |
914 | (*) Control dependencies pair normally with other types of barriers. |
915 | ||
f1ab25a3 PM |
916 | (*) Control dependencies do -not- provide multicopy atomicity. If you |
917 | need all the CPUs to see a given store at the same time, use smp_mb(). | |
108b42b4 | 918 | |
c8241f85 PM |
919 | (*) Compilers do not understand control dependencies. It is therefore |
920 | your job to ensure that they do not break your code. | |
921 | ||
108b42b4 DH |
922 | |
923 | SMP BARRIER PAIRING | |
924 | ------------------- | |
925 | ||
926 | When dealing with CPU-CPU interactions, certain types of memory barrier should | |
927 | always be paired. A lack of appropriate pairing is almost certainly an error. | |
928 | ||
ff382810 | 929 | General barriers pair with each other, though they also pair with most |
f1ab25a3 PM |
930 | other types of barriers, albeit without multicopy atomicity. An acquire |
931 | barrier pairs with a release barrier, but both may also pair with other | |
932 | barriers, including of course general barriers. A write barrier pairs | |
933 | with a data dependency barrier, a control dependency, an acquire barrier, | |
934 | a release barrier, a read barrier, or a general barrier. Similarly a | |
935 | read barrier, control dependency, or a data dependency barrier pairs | |
936 | with a write barrier, an acquire barrier, a release barrier, or a | |
937 | general barrier: | |
108b42b4 | 938 | |
2ecf8101 PM |
939 | CPU 1 CPU 2 |
940 | =============== =============== | |
9af194ce | 941 | WRITE_ONCE(a, 1); |
108b42b4 | 942 | <write barrier> |
9af194ce | 943 | WRITE_ONCE(b, 2); x = READ_ONCE(b); |
2ecf8101 | 944 | <read barrier> |
9af194ce | 945 | y = READ_ONCE(a); |
108b42b4 DH |
946 | |
947 | Or: | |
948 | ||
2ecf8101 PM |
949 | CPU 1 CPU 2 |
950 | =============== =============================== | |
108b42b4 DH |
951 | a = 1; |
952 | <write barrier> | |
9af194ce | 953 | WRITE_ONCE(b, &a); x = READ_ONCE(b); |
2ecf8101 PM |
954 | <data dependency barrier> |
955 | y = *x; | |
108b42b4 | 956 | |
ff382810 PM |
957 | Or even: |
958 | ||
959 | CPU 1 CPU 2 | |
960 | =============== =============================== | |
9af194ce | 961 | r1 = READ_ONCE(y); |
ff382810 | 962 | <general barrier> |
d92f842b | 963 | WRITE_ONCE(x, 1); if (r2 = READ_ONCE(x)) { |
ff382810 | 964 | <implicit control dependency> |
9af194ce | 965 | WRITE_ONCE(y, 1); |
ff382810 PM |
966 | } |
967 | ||
968 | assert(r1 == 0 || r2 == 0); | |
969 | ||
108b42b4 DH |
970 | Basically, the read barrier always has to be there, even though it can be of |
971 | the "weaker" type. | |
972 | ||
670bd95e | 973 | [!] Note that the stores before the write barrier would normally be expected to |
81fc6323 | 974 | match the loads after the read barrier or the data dependency barrier, and vice |
670bd95e DH |
975 | versa: |
976 | ||
2ecf8101 PM |
977 | CPU 1 CPU 2 |
978 | =================== =================== | |
9af194ce PM |
979 | WRITE_ONCE(a, 1); }---- --->{ v = READ_ONCE(c); |
980 | WRITE_ONCE(b, 2); } \ / { w = READ_ONCE(d); | |
2ecf8101 | 981 | <write barrier> \ <read barrier> |
9af194ce PM |
982 | WRITE_ONCE(c, 3); } / \ { x = READ_ONCE(a); |
983 | WRITE_ONCE(d, 4); }---- --->{ y = READ_ONCE(b); | |
670bd95e | 984 | |
108b42b4 DH |
985 | |
986 | EXAMPLES OF MEMORY BARRIER SEQUENCES | |
987 | ------------------------------------ | |
988 | ||
81fc6323 | 989 | Firstly, write barriers act as partial orderings on store operations. |
108b42b4 DH |
990 | Consider the following sequence of events: |
991 | ||
992 | CPU 1 | |
993 | ======================= | |
994 | STORE A = 1 | |
995 | STORE B = 2 | |
996 | STORE C = 3 | |
997 | <write barrier> | |
998 | STORE D = 4 | |
999 | STORE E = 5 | |
1000 | ||
1001 | This sequence of events is committed to the memory coherence system in an order | |
1002 | that the rest of the system might perceive as the unordered set of { STORE A, | |
80f7228b | 1003 | STORE B, STORE C } all occurring before the unordered set of { STORE D, STORE E |
108b42b4 DH |
1004 | }: |
1005 | ||
1006 | +-------+ : : | |
1007 | | | +------+ | |
1008 | | |------>| C=3 | } /\ | |
81fc6323 JP |
1009 | | | : +------+ }----- \ -----> Events perceptible to |
1010 | | | : | A=1 | } \/ the rest of the system | |
108b42b4 DH |
1011 | | | : +------+ } |
1012 | | CPU 1 | : | B=2 | } | |
1013 | | | +------+ } | |
1014 | | | wwwwwwwwwwwwwwww } <--- At this point the write barrier | |
1015 | | | +------+ } requires all stores prior to the | |
1016 | | | : | E=5 | } barrier to be committed before | |
81fc6323 | 1017 | | | : +------+ } further stores may take place |
108b42b4 DH |
1018 | | |------>| D=4 | } |
1019 | | | +------+ | |
1020 | +-------+ : : | |
1021 | | | |
670bd95e DH |
1022 | | Sequence in which stores are committed to the |
1023 | | memory system by CPU 1 | |
108b42b4 DH |
1024 | V |
1025 | ||
1026 | ||
81fc6323 | 1027 | Secondly, data dependency barriers act as partial orderings on data-dependent |
108b42b4 DH |
1028 | loads. Consider the following sequence of events: |
1029 | ||
1030 | CPU 1 CPU 2 | |
1031 | ======================= ======================= | |
c14038c3 | 1032 | { B = 7; X = 9; Y = 8; C = &Y } |
108b42b4 DH |
1033 | STORE A = 1 |
1034 | STORE B = 2 | |
1035 | <write barrier> | |
1036 | STORE C = &B LOAD X | |
1037 | STORE D = 4 LOAD C (gets &B) | |
1038 | LOAD *C (reads B) | |
1039 | ||
1040 | Without intervention, CPU 2 may perceive the events on CPU 1 in some | |
1041 | effectively random order, despite the write barrier issued by CPU 1: | |
1042 | ||
1043 | +-------+ : : : : | |
1044 | | | +------+ +-------+ | Sequence of update | |
1045 | | |------>| B=2 |----- --->| Y->8 | | of perception on | |
1046 | | | : +------+ \ +-------+ | CPU 2 | |
1047 | | CPU 1 | : | A=1 | \ --->| C->&Y | V | |
1048 | | | +------+ | +-------+ | |
1049 | | | wwwwwwwwwwwwwwww | : : | |
1050 | | | +------+ | : : | |
1051 | | | : | C=&B |--- | : : +-------+ | |
1052 | | | : +------+ \ | +-------+ | | | |
1053 | | |------>| D=4 | ----------->| C->&B |------>| | | |
1054 | | | +------+ | +-------+ | | | |
1055 | +-------+ : : | : : | | | |
1056 | | : : | | | |
1057 | | : : | CPU 2 | | |
1058 | | +-------+ | | | |
1059 | Apparently incorrect ---> | | B->7 |------>| | | |
1060 | perception of B (!) | +-------+ | | | |
1061 | | : : | | | |
1062 | | +-------+ | | | |
1063 | The load of X holds ---> \ | X->9 |------>| | | |
1064 | up the maintenance \ +-------+ | | | |
1065 | of coherence of B ----->| B->2 | +-------+ | |
1066 | +-------+ | |
1067 | : : | |
1068 | ||
1069 | ||
1070 | In the above example, CPU 2 perceives that B is 7, despite the load of *C | |
670e9f34 | 1071 | (which would be B) coming after the LOAD of C. |
108b42b4 DH |
1072 | |
1073 | If, however, a data dependency barrier were to be placed between the load of C | |
c14038c3 DH |
1074 | and the load of *C (ie: B) on CPU 2: |
1075 | ||
1076 | CPU 1 CPU 2 | |
1077 | ======================= ======================= | |
1078 | { B = 7; X = 9; Y = 8; C = &Y } | |
1079 | STORE A = 1 | |
1080 | STORE B = 2 | |
1081 | <write barrier> | |
1082 | STORE C = &B LOAD X | |
1083 | STORE D = 4 LOAD C (gets &B) | |
1084 | <data dependency barrier> | |
1085 | LOAD *C (reads B) | |
1086 | ||
1087 | then the following will occur: | |
108b42b4 DH |
1088 | |
1089 | +-------+ : : : : | |
1090 | | | +------+ +-------+ | |
1091 | | |------>| B=2 |----- --->| Y->8 | | |
1092 | | | : +------+ \ +-------+ | |
1093 | | CPU 1 | : | A=1 | \ --->| C->&Y | | |
1094 | | | +------+ | +-------+ | |
1095 | | | wwwwwwwwwwwwwwww | : : | |
1096 | | | +------+ | : : | |
1097 | | | : | C=&B |--- | : : +-------+ | |
1098 | | | : +------+ \ | +-------+ | | | |
1099 | | |------>| D=4 | ----------->| C->&B |------>| | | |
1100 | | | +------+ | +-------+ | | | |
1101 | +-------+ : : | : : | | | |
1102 | | : : | | | |
1103 | | : : | CPU 2 | | |
1104 | | +-------+ | | | |
670bd95e DH |
1105 | | | X->9 |------>| | |
1106 | | +-------+ | | | |
1107 | Makes sure all effects ---> \ ddddddddddddddddd | | | |
1108 | prior to the store of C \ +-------+ | | | |
1109 | are perceptible to ----->| B->2 |------>| | | |
1110 | subsequent loads +-------+ | | | |
108b42b4 DH |
1111 | : : +-------+ |
1112 | ||
1113 | ||
1114 | And thirdly, a read barrier acts as a partial order on loads. Consider the | |
1115 | following sequence of events: | |
1116 | ||
1117 | CPU 1 CPU 2 | |
1118 | ======================= ======================= | |
670bd95e | 1119 | { A = 0, B = 9 } |
108b42b4 | 1120 | STORE A=1 |
108b42b4 | 1121 | <write barrier> |
670bd95e | 1122 | STORE B=2 |
108b42b4 | 1123 | LOAD B |
670bd95e | 1124 | LOAD A |
108b42b4 DH |
1125 | |
1126 | Without intervention, CPU 2 may then choose to perceive the events on CPU 1 in | |
1127 | some effectively random order, despite the write barrier issued by CPU 1: | |
1128 | ||
670bd95e DH |
1129 | +-------+ : : : : |
1130 | | | +------+ +-------+ | |
1131 | | |------>| A=1 |------ --->| A->0 | | |
1132 | | | +------+ \ +-------+ | |
1133 | | CPU 1 | wwwwwwwwwwwwwwww \ --->| B->9 | | |
1134 | | | +------+ | +-------+ | |
1135 | | |------>| B=2 |--- | : : | |
1136 | | | +------+ \ | : : +-------+ | |
1137 | +-------+ : : \ | +-------+ | | | |
1138 | ---------->| B->2 |------>| | | |
1139 | | +-------+ | CPU 2 | | |
1140 | | | A->0 |------>| | | |
1141 | | +-------+ | | | |
1142 | | : : +-------+ | |
1143 | \ : : | |
1144 | \ +-------+ | |
1145 | ---->| A->1 | | |
1146 | +-------+ | |
1147 | : : | |
108b42b4 | 1148 | |
670bd95e | 1149 | |
6bc39274 | 1150 | If, however, a read barrier were to be placed between the load of B and the |
670bd95e DH |
1151 | load of A on CPU 2: |
1152 | ||
1153 | CPU 1 CPU 2 | |
1154 | ======================= ======================= | |
1155 | { A = 0, B = 9 } | |
1156 | STORE A=1 | |
1157 | <write barrier> | |
1158 | STORE B=2 | |
1159 | LOAD B | |
1160 | <read barrier> | |
1161 | LOAD A | |
1162 | ||
1163 | then the partial ordering imposed by CPU 1 will be perceived correctly by CPU | |
1164 | 2: | |
1165 | ||
1166 | +-------+ : : : : | |
1167 | | | +------+ +-------+ | |
1168 | | |------>| A=1 |------ --->| A->0 | | |
1169 | | | +------+ \ +-------+ | |
1170 | | CPU 1 | wwwwwwwwwwwwwwww \ --->| B->9 | | |
1171 | | | +------+ | +-------+ | |
1172 | | |------>| B=2 |--- | : : | |
1173 | | | +------+ \ | : : +-------+ | |
1174 | +-------+ : : \ | +-------+ | | | |
1175 | ---------->| B->2 |------>| | | |
1176 | | +-------+ | CPU 2 | | |
1177 | | : : | | | |
1178 | | : : | | | |
1179 | At this point the read ----> \ rrrrrrrrrrrrrrrrr | | | |
1180 | barrier causes all effects \ +-------+ | | | |
1181 | prior to the storage of B ---->| A->1 |------>| | | |
1182 | to be perceptible to CPU 2 +-------+ | | | |
1183 | : : +-------+ | |
1184 | ||
1185 | ||
1186 | To illustrate this more completely, consider what could happen if the code | |
1187 | contained a load of A either side of the read barrier: | |
1188 | ||
1189 | CPU 1 CPU 2 | |
1190 | ======================= ======================= | |
1191 | { A = 0, B = 9 } | |
1192 | STORE A=1 | |
1193 | <write barrier> | |
1194 | STORE B=2 | |
1195 | LOAD B | |
1196 | LOAD A [first load of A] | |
1197 | <read barrier> | |
1198 | LOAD A [second load of A] | |
1199 | ||
1200 | Even though the two loads of A both occur after the load of B, they may both | |
1201 | come up with different values: | |
1202 | ||
1203 | +-------+ : : : : | |
1204 | | | +------+ +-------+ | |
1205 | | |------>| A=1 |------ --->| A->0 | | |
1206 | | | +------+ \ +-------+ | |
1207 | | CPU 1 | wwwwwwwwwwwwwwww \ --->| B->9 | | |
1208 | | | +------+ | +-------+ | |
1209 | | |------>| B=2 |--- | : : | |
1210 | | | +------+ \ | : : +-------+ | |
1211 | +-------+ : : \ | +-------+ | | | |
1212 | ---------->| B->2 |------>| | | |
1213 | | +-------+ | CPU 2 | | |
1214 | | : : | | | |
1215 | | : : | | | |
1216 | | +-------+ | | | |
1217 | | | A->0 |------>| 1st | | |
1218 | | +-------+ | | | |
1219 | At this point the read ----> \ rrrrrrrrrrrrrrrrr | | | |
1220 | barrier causes all effects \ +-------+ | | | |
1221 | prior to the storage of B ---->| A->1 |------>| 2nd | | |
1222 | to be perceptible to CPU 2 +-------+ | | | |
1223 | : : +-------+ | |
1224 | ||
1225 | ||
1226 | But it may be that the update to A from CPU 1 becomes perceptible to CPU 2 | |
1227 | before the read barrier completes anyway: | |
1228 | ||
1229 | +-------+ : : : : | |
1230 | | | +------+ +-------+ | |
1231 | | |------>| A=1 |------ --->| A->0 | | |
1232 | | | +------+ \ +-------+ | |
1233 | | CPU 1 | wwwwwwwwwwwwwwww \ --->| B->9 | | |
1234 | | | +------+ | +-------+ | |
1235 | | |------>| B=2 |--- | : : | |
1236 | | | +------+ \ | : : +-------+ | |
1237 | +-------+ : : \ | +-------+ | | | |
1238 | ---------->| B->2 |------>| | | |
1239 | | +-------+ | CPU 2 | | |
1240 | | : : | | | |
1241 | \ : : | | | |
1242 | \ +-------+ | | | |
1243 | ---->| A->1 |------>| 1st | | |
1244 | +-------+ | | | |
1245 | rrrrrrrrrrrrrrrrr | | | |
1246 | +-------+ | | | |
1247 | | A->1 |------>| 2nd | | |
1248 | +-------+ | | | |
1249 | : : +-------+ | |
1250 | ||
1251 | ||
1252 | The guarantee is that the second load will always come up with A == 1 if the | |
1253 | load of B came up with B == 2. No such guarantee exists for the first load of | |
1254 | A; that may come up with either A == 0 or A == 1. | |
1255 | ||
1256 | ||
1257 | READ MEMORY BARRIERS VS LOAD SPECULATION | |
1258 | ---------------------------------------- | |
1259 | ||
1260 | Many CPUs speculate with loads: that is they see that they will need to load an | |
1261 | item from memory, and they find a time where they're not using the bus for any | |
1262 | other loads, and so do the load in advance - even though they haven't actually | |
1263 | got to that point in the instruction execution flow yet. This permits the | |
1264 | actual load instruction to potentially complete immediately because the CPU | |
1265 | already has the value to hand. | |
1266 | ||
1267 | It may turn out that the CPU didn't actually need the value - perhaps because a | |
1268 | branch circumvented the load - in which case it can discard the value or just | |
1269 | cache it for later use. | |
1270 | ||
1271 | Consider: | |
1272 | ||
e0edc78f | 1273 | CPU 1 CPU 2 |
670bd95e | 1274 | ======================= ======================= |
e0edc78f IM |
1275 | LOAD B |
1276 | DIVIDE } Divide instructions generally | |
1277 | DIVIDE } take a long time to perform | |
1278 | LOAD A | |
670bd95e DH |
1279 | |
1280 | Which might appear as this: | |
1281 | ||
1282 | : : +-------+ | |
1283 | +-------+ | | | |
1284 | --->| B->2 |------>| | | |
1285 | +-------+ | CPU 2 | | |
1286 | : :DIVIDE | | | |
1287 | +-------+ | | | |
1288 | The CPU being busy doing a ---> --->| A->0 |~~~~ | | | |
1289 | division speculates on the +-------+ ~ | | | |
1290 | LOAD of A : : ~ | | | |
1291 | : :DIVIDE | | | |
1292 | : : ~ | | | |
1293 | Once the divisions are complete --> : : ~-->| | | |
1294 | the CPU can then perform the : : | | | |
1295 | LOAD with immediate effect : : +-------+ | |
1296 | ||
1297 | ||
1298 | Placing a read barrier or a data dependency barrier just before the second | |
1299 | load: | |
1300 | ||
e0edc78f | 1301 | CPU 1 CPU 2 |
670bd95e | 1302 | ======================= ======================= |
e0edc78f IM |
1303 | LOAD B |
1304 | DIVIDE | |
1305 | DIVIDE | |
670bd95e | 1306 | <read barrier> |
e0edc78f | 1307 | LOAD A |
670bd95e DH |
1308 | |
1309 | will force any value speculatively obtained to be reconsidered to an extent | |
1310 | dependent on the type of barrier used. If there was no change made to the | |
1311 | speculated memory location, then the speculated value will just be used: | |
1312 | ||
1313 | : : +-------+ | |
1314 | +-------+ | | | |
1315 | --->| B->2 |------>| | | |
1316 | +-------+ | CPU 2 | | |
1317 | : :DIVIDE | | | |
1318 | +-------+ | | | |
1319 | The CPU being busy doing a ---> --->| A->0 |~~~~ | | | |
1320 | division speculates on the +-------+ ~ | | | |
1321 | LOAD of A : : ~ | | | |
1322 | : :DIVIDE | | | |
1323 | : : ~ | | | |
1324 | : : ~ | | | |
1325 | rrrrrrrrrrrrrrrr~ | | | |
1326 | : : ~ | | | |
1327 | : : ~-->| | | |
1328 | : : | | | |
1329 | : : +-------+ | |
1330 | ||
1331 | ||
1332 | but if there was an update or an invalidation from another CPU pending, then | |
1333 | the speculation will be cancelled and the value reloaded: | |
1334 | ||
1335 | : : +-------+ | |
1336 | +-------+ | | | |
1337 | --->| B->2 |------>| | | |
1338 | +-------+ | CPU 2 | | |
1339 | : :DIVIDE | | | |
1340 | +-------+ | | | |
1341 | The CPU being busy doing a ---> --->| A->0 |~~~~ | | | |
1342 | division speculates on the +-------+ ~ | | | |
1343 | LOAD of A : : ~ | | | |
1344 | : :DIVIDE | | | |
1345 | : : ~ | | | |
1346 | : : ~ | | | |
1347 | rrrrrrrrrrrrrrrrr | | | |
1348 | +-------+ | | | |
1349 | The speculation is discarded ---> --->| A->1 |------>| | | |
1350 | and an updated value is +-------+ | | | |
1351 | retrieved : : +-------+ | |
108b42b4 DH |
1352 | |
1353 | ||
f1ab25a3 PM |
1354 | MULTICOPY ATOMICITY |
1355 | -------------------- | |
1356 | ||
1357 | Multicopy atomicity is a deeply intuitive notion about ordering that is | |
1358 | not always provided by real computer systems, namely that a given store | |
0902b1f4 AS |
1359 | becomes visible at the same time to all CPUs, or, alternatively, that all |
1360 | CPUs agree on the order in which all stores become visible. However, | |
1361 | support of full multicopy atomicity would rule out valuable hardware | |
1362 | optimizations, so a weaker form called ``other multicopy atomicity'' | |
1363 | instead guarantees only that a given store becomes visible at the same | |
1364 | time to all -other- CPUs. The remainder of this document discusses this | |
1365 | weaker form, but for brevity will call it simply ``multicopy atomicity''. | |
241e6663 | 1366 | |
f1ab25a3 | 1367 | The following example demonstrates multicopy atomicity: |
241e6663 PM |
1368 | |
1369 | CPU 1 CPU 2 CPU 3 | |
1370 | ======================= ======================= ======================= | |
1371 | { X = 0, Y = 0 } | |
f1ab25a3 PM |
1372 | STORE X=1 r1=LOAD X (reads 1) LOAD Y (reads 1) |
1373 | <general barrier> <read barrier> | |
1374 | STORE Y=r1 LOAD X | |
241e6663 | 1375 | |
0902b1f4 AS |
1376 | Suppose that CPU 2's load from X returns 1, which it then stores to Y, |
1377 | and CPU 3's load from Y returns 1. This indicates that CPU 1's store | |
1378 | to X precedes CPU 2's load from X and that CPU 2's store to Y precedes | |
1379 | CPU 3's load from Y. In addition, the memory barriers guarantee that | |
1380 | CPU 2 executes its load before its store, and CPU 3 loads from Y before | |
1381 | it loads from X. The question is then "Can CPU 3's load from X return 0?" | |
241e6663 | 1382 | |
0902b1f4 | 1383 | Because CPU 3's load from X in some sense comes after CPU 2's load, it |
241e6663 | 1384 | is natural to expect that CPU 3's load from X must therefore return 1. |
0902b1f4 AS |
1385 | This expectation follows from multicopy atomicity: if a load executing |
1386 | on CPU B follows a load from the same variable executing on CPU A (and | |
1387 | CPU A did not originally store the value which it read), then on | |
1388 | multicopy-atomic systems, CPU B's load must return either the same value | |
1389 | that CPU A's load did or some later value. However, the Linux kernel | |
1390 | does not require systems to be multicopy atomic. | |
1391 | ||
1392 | The use of a general memory barrier in the example above compensates | |
1393 | for any lack of multicopy atomicity. In the example, if CPU 2's load | |
1394 | from X returns 1 and CPU 3's load from Y returns 1, then CPU 3's load | |
1395 | from X must indeed also return 1. | |
f1ab25a3 PM |
1396 | |
1397 | However, dependencies, read barriers, and write barriers are not always | |
1398 | able to compensate for non-multicopy atomicity. For example, suppose | |
1399 | that CPU 2's general barrier is removed from the above example, leaving | |
1400 | only the data dependency shown below: | |
241e6663 PM |
1401 | |
1402 | CPU 1 CPU 2 CPU 3 | |
1403 | ======================= ======================= ======================= | |
1404 | { X = 0, Y = 0 } | |
f1ab25a3 PM |
1405 | STORE X=1 r1=LOAD X (reads 1) LOAD Y (reads 1) |
1406 | <data dependency> <read barrier> | |
1407 | STORE Y=r1 LOAD X (reads 0) | |
1408 | ||
1409 | This substitution allows non-multicopy atomicity to run rampant: in | |
1410 | this example, it is perfectly legal for CPU 2's load from X to return 1, | |
1411 | CPU 3's load from Y to return 1, and its load from X to return 0. | |
1412 | ||
1413 | The key point is that although CPU 2's data dependency orders its load | |
0902b1f4 AS |
1414 | and store, it does not guarantee to order CPU 1's store. Thus, if this |
1415 | example runs on a non-multicopy-atomic system where CPUs 1 and 2 share a | |
1416 | store buffer or a level of cache, CPU 2 might have early access to CPU 1's | |
1417 | writes. General barriers are therefore required to ensure that all CPUs | |
1418 | agree on the combined order of multiple accesses. | |
f1ab25a3 PM |
1419 | |
1420 | General barriers can compensate not only for non-multicopy atomicity, | |
1421 | but can also generate additional ordering that can ensure that -all- | |
1422 | CPUs will perceive the same order of -all- operations. In contrast, a | |
1423 | chain of release-acquire pairs do not provide this additional ordering, | |
1424 | which means that only those CPUs on the chain are guaranteed to agree | |
1425 | on the combined order of the accesses. For example, switching to C code | |
1426 | in deference to the ghost of Herman Hollerith: | |
c535cc92 PM |
1427 | |
1428 | int u, v, x, y, z; | |
1429 | ||
1430 | void cpu0(void) | |
1431 | { | |
1432 | r0 = smp_load_acquire(&x); | |
1433 | WRITE_ONCE(u, 1); | |
1434 | smp_store_release(&y, 1); | |
1435 | } | |
1436 | ||
1437 | void cpu1(void) | |
1438 | { | |
1439 | r1 = smp_load_acquire(&y); | |
1440 | r4 = READ_ONCE(v); | |
1441 | r5 = READ_ONCE(u); | |
1442 | smp_store_release(&z, 1); | |
1443 | } | |
1444 | ||
1445 | void cpu2(void) | |
1446 | { | |
1447 | r2 = smp_load_acquire(&z); | |
1448 | smp_store_release(&x, 1); | |
1449 | } | |
1450 | ||
1451 | void cpu3(void) | |
1452 | { | |
1453 | WRITE_ONCE(v, 1); | |
1454 | smp_mb(); | |
1455 | r3 = READ_ONCE(u); | |
1456 | } | |
1457 | ||
f1ab25a3 PM |
1458 | Because cpu0(), cpu1(), and cpu2() participate in a chain of |
1459 | smp_store_release()/smp_load_acquire() pairs, the following outcome | |
1460 | is prohibited: | |
c535cc92 PM |
1461 | |
1462 | r0 == 1 && r1 == 1 && r2 == 1 | |
1463 | ||
1464 | Furthermore, because of the release-acquire relationship between cpu0() | |
1465 | and cpu1(), cpu1() must see cpu0()'s writes, so that the following | |
1466 | outcome is prohibited: | |
1467 | ||
1468 | r1 == 1 && r5 == 0 | |
1469 | ||
f1ab25a3 PM |
1470 | However, the ordering provided by a release-acquire chain is local |
1471 | to the CPUs participating in that chain and does not apply to cpu3(), | |
1472 | at least aside from stores. Therefore, the following outcome is possible: | |
c535cc92 PM |
1473 | |
1474 | r0 == 0 && r1 == 1 && r2 == 1 && r3 == 0 && r4 == 0 | |
1475 | ||
37ef0341 PM |
1476 | As an aside, the following outcome is also possible: |
1477 | ||
1478 | r0 == 0 && r1 == 1 && r2 == 1 && r3 == 0 && r4 == 0 && r5 == 1 | |
1479 | ||
c535cc92 PM |
1480 | Although cpu0(), cpu1(), and cpu2() will see their respective reads and |
1481 | writes in order, CPUs not involved in the release-acquire chain might | |
1482 | well disagree on the order. This disagreement stems from the fact that | |
1483 | the weak memory-barrier instructions used to implement smp_load_acquire() | |
1484 | and smp_store_release() are not required to order prior stores against | |
1485 | subsequent loads in all cases. This means that cpu3() can see cpu0()'s | |
1486 | store to u as happening -after- cpu1()'s load from v, even though | |
1487 | both cpu0() and cpu1() agree that these two operations occurred in the | |
1488 | intended order. | |
1489 | ||
1490 | However, please keep in mind that smp_load_acquire() is not magic. | |
1491 | In particular, it simply reads from its argument with ordering. It does | |
1492 | -not- ensure that any particular value will be read. Therefore, the | |
1493 | following outcome is possible: | |
1494 | ||
1495 | r0 == 0 && r1 == 0 && r2 == 0 && r5 == 0 | |
1496 | ||
1497 | Note that this outcome can happen even on a mythical sequentially | |
1498 | consistent system where nothing is ever reordered. | |
1499 | ||
f1ab25a3 PM |
1500 | To reiterate, if your code requires full ordering of all operations, |
1501 | use general barriers throughout. | |
241e6663 PM |
1502 | |
1503 | ||
108b42b4 DH |
1504 | ======================== |
1505 | EXPLICIT KERNEL BARRIERS | |
1506 | ======================== | |
1507 | ||
1508 | The Linux kernel has a variety of different barriers that act at different | |
1509 | levels: | |
1510 | ||
1511 | (*) Compiler barrier. | |
1512 | ||
1513 | (*) CPU memory barriers. | |
1514 | ||
1515 | (*) MMIO write barrier. | |
1516 | ||
1517 | ||
1518 | COMPILER BARRIER | |
1519 | ---------------- | |
1520 | ||
1521 | The Linux kernel has an explicit compiler barrier function that prevents the | |
1522 | compiler from moving the memory accesses either side of it to the other side: | |
1523 | ||
1524 | barrier(); | |
1525 | ||
9af194ce PM |
1526 | This is a general barrier -- there are no read-read or write-write |
1527 | variants of barrier(). However, READ_ONCE() and WRITE_ONCE() can be | |
1528 | thought of as weak forms of barrier() that affect only the specific | |
1529 | accesses flagged by the READ_ONCE() or WRITE_ONCE(). | |
108b42b4 | 1530 | |
692118da PM |
1531 | The barrier() function has the following effects: |
1532 | ||
1533 | (*) Prevents the compiler from reordering accesses following the | |
1534 | barrier() to precede any accesses preceding the barrier(). | |
1535 | One example use for this property is to ease communication between | |
1536 | interrupt-handler code and the code that was interrupted. | |
1537 | ||
1538 | (*) Within a loop, forces the compiler to load the variables used | |
1539 | in that loop's conditional on each pass through that loop. | |
1540 | ||
9af194ce PM |
1541 | The READ_ONCE() and WRITE_ONCE() functions can prevent any number of |
1542 | optimizations that, while perfectly safe in single-threaded code, can | |
1543 | be fatal in concurrent code. Here are some examples of these sorts | |
1544 | of optimizations: | |
692118da | 1545 | |
449f7413 PM |
1546 | (*) The compiler is within its rights to reorder loads and stores |
1547 | to the same variable, and in some cases, the CPU is within its | |
1548 | rights to reorder loads to the same variable. This means that | |
1549 | the following code: | |
1550 | ||
1551 | a[0] = x; | |
1552 | a[1] = x; | |
1553 | ||
1554 | Might result in an older value of x stored in a[1] than in a[0]. | |
1555 | Prevent both the compiler and the CPU from doing this as follows: | |
1556 | ||
9af194ce PM |
1557 | a[0] = READ_ONCE(x); |
1558 | a[1] = READ_ONCE(x); | |
449f7413 | 1559 | |
9af194ce PM |
1560 | In short, READ_ONCE() and WRITE_ONCE() provide cache coherence for |
1561 | accesses from multiple CPUs to a single variable. | |
449f7413 | 1562 | |
692118da PM |
1563 | (*) The compiler is within its rights to merge successive loads from |
1564 | the same variable. Such merging can cause the compiler to "optimize" | |
1565 | the following code: | |
1566 | ||
1567 | while (tmp = a) | |
1568 | do_something_with(tmp); | |
1569 | ||
1570 | into the following code, which, although in some sense legitimate | |
1571 | for single-threaded code, is almost certainly not what the developer | |
1572 | intended: | |
1573 | ||
1574 | if (tmp = a) | |
1575 | for (;;) | |
1576 | do_something_with(tmp); | |
1577 | ||
9af194ce | 1578 | Use READ_ONCE() to prevent the compiler from doing this to you: |
692118da | 1579 | |
9af194ce | 1580 | while (tmp = READ_ONCE(a)) |
692118da PM |
1581 | do_something_with(tmp); |
1582 | ||
1583 | (*) The compiler is within its rights to reload a variable, for example, | |
1584 | in cases where high register pressure prevents the compiler from | |
1585 | keeping all data of interest in registers. The compiler might | |
1586 | therefore optimize the variable 'tmp' out of our previous example: | |
1587 | ||
1588 | while (tmp = a) | |
1589 | do_something_with(tmp); | |
1590 | ||
1591 | This could result in the following code, which is perfectly safe in | |
1592 | single-threaded code, but can be fatal in concurrent code: | |
1593 | ||
1594 | while (a) | |
1595 | do_something_with(a); | |
1596 | ||
1597 | For example, the optimized version of this code could result in | |
1598 | passing a zero to do_something_with() in the case where the variable | |
1599 | a was modified by some other CPU between the "while" statement and | |
1600 | the call to do_something_with(). | |
1601 | ||
9af194ce | 1602 | Again, use READ_ONCE() to prevent the compiler from doing this: |
692118da | 1603 | |
9af194ce | 1604 | while (tmp = READ_ONCE(a)) |
692118da PM |
1605 | do_something_with(tmp); |
1606 | ||
1607 | Note that if the compiler runs short of registers, it might save | |
1608 | tmp onto the stack. The overhead of this saving and later restoring | |
1609 | is why compilers reload variables. Doing so is perfectly safe for | |
1610 | single-threaded code, so you need to tell the compiler about cases | |
1611 | where it is not safe. | |
1612 | ||
1613 | (*) The compiler is within its rights to omit a load entirely if it knows | |
1614 | what the value will be. For example, if the compiler can prove that | |
1615 | the value of variable 'a' is always zero, it can optimize this code: | |
1616 | ||
1617 | while (tmp = a) | |
1618 | do_something_with(tmp); | |
1619 | ||
1620 | Into this: | |
1621 | ||
1622 | do { } while (0); | |
1623 | ||
9af194ce PM |
1624 | This transformation is a win for single-threaded code because it |
1625 | gets rid of a load and a branch. The problem is that the compiler | |
1626 | will carry out its proof assuming that the current CPU is the only | |
1627 | one updating variable 'a'. If variable 'a' is shared, then the | |
1628 | compiler's proof will be erroneous. Use READ_ONCE() to tell the | |
1629 | compiler that it doesn't know as much as it thinks it does: | |
692118da | 1630 | |
9af194ce | 1631 | while (tmp = READ_ONCE(a)) |
692118da PM |
1632 | do_something_with(tmp); |
1633 | ||
1634 | But please note that the compiler is also closely watching what you | |
9af194ce | 1635 | do with the value after the READ_ONCE(). For example, suppose you |
692118da PM |
1636 | do the following and MAX is a preprocessor macro with the value 1: |
1637 | ||
9af194ce | 1638 | while ((tmp = READ_ONCE(a)) % MAX) |
692118da PM |
1639 | do_something_with(tmp); |
1640 | ||
1641 | Then the compiler knows that the result of the "%" operator applied | |
1642 | to MAX will always be zero, again allowing the compiler to optimize | |
1643 | the code into near-nonexistence. (It will still load from the | |
1644 | variable 'a'.) | |
1645 | ||
1646 | (*) Similarly, the compiler is within its rights to omit a store entirely | |
1647 | if it knows that the variable already has the value being stored. | |
1648 | Again, the compiler assumes that the current CPU is the only one | |
1649 | storing into the variable, which can cause the compiler to do the | |
1650 | wrong thing for shared variables. For example, suppose you have | |
1651 | the following: | |
1652 | ||
1653 | a = 0; | |
65f95ff2 | 1654 | ... Code that does not store to variable a ... |
692118da PM |
1655 | a = 0; |
1656 | ||
1657 | The compiler sees that the value of variable 'a' is already zero, so | |
1658 | it might well omit the second store. This would come as a fatal | |
1659 | surprise if some other CPU might have stored to variable 'a' in the | |
1660 | meantime. | |
1661 | ||
9af194ce | 1662 | Use WRITE_ONCE() to prevent the compiler from making this sort of |
692118da PM |
1663 | wrong guess: |
1664 | ||
9af194ce | 1665 | WRITE_ONCE(a, 0); |
65f95ff2 | 1666 | ... Code that does not store to variable a ... |
9af194ce | 1667 | WRITE_ONCE(a, 0); |
692118da PM |
1668 | |
1669 | (*) The compiler is within its rights to reorder memory accesses unless | |
1670 | you tell it not to. For example, consider the following interaction | |
1671 | between process-level code and an interrupt handler: | |
1672 | ||
1673 | void process_level(void) | |
1674 | { | |
1675 | msg = get_message(); | |
1676 | flag = true; | |
1677 | } | |
1678 | ||
1679 | void interrupt_handler(void) | |
1680 | { | |
1681 | if (flag) | |
1682 | process_message(msg); | |
1683 | } | |
1684 | ||
df5cbb27 | 1685 | There is nothing to prevent the compiler from transforming |
692118da PM |
1686 | process_level() to the following, in fact, this might well be a |
1687 | win for single-threaded code: | |
1688 | ||
1689 | void process_level(void) | |
1690 | { | |
1691 | flag = true; | |
1692 | msg = get_message(); | |
1693 | } | |
1694 | ||
1695 | If the interrupt occurs between these two statement, then | |
9af194ce | 1696 | interrupt_handler() might be passed a garbled msg. Use WRITE_ONCE() |
692118da PM |
1697 | to prevent this as follows: |
1698 | ||
1699 | void process_level(void) | |
1700 | { | |
9af194ce PM |
1701 | WRITE_ONCE(msg, get_message()); |
1702 | WRITE_ONCE(flag, true); | |
692118da PM |
1703 | } |
1704 | ||
1705 | void interrupt_handler(void) | |
1706 | { | |
9af194ce PM |
1707 | if (READ_ONCE(flag)) |
1708 | process_message(READ_ONCE(msg)); | |
692118da PM |
1709 | } |
1710 | ||
9af194ce PM |
1711 | Note that the READ_ONCE() and WRITE_ONCE() wrappers in |
1712 | interrupt_handler() are needed if this interrupt handler can itself | |
1713 | be interrupted by something that also accesses 'flag' and 'msg', | |
1714 | for example, a nested interrupt or an NMI. Otherwise, READ_ONCE() | |
1715 | and WRITE_ONCE() are not needed in interrupt_handler() other than | |
1716 | for documentation purposes. (Note also that nested interrupts | |
1717 | do not typically occur in modern Linux kernels, in fact, if an | |
1718 | interrupt handler returns with interrupts enabled, you will get a | |
1719 | WARN_ONCE() splat.) | |
1720 | ||
1721 | You should assume that the compiler can move READ_ONCE() and | |
1722 | WRITE_ONCE() past code not containing READ_ONCE(), WRITE_ONCE(), | |
1723 | barrier(), or similar primitives. | |
1724 | ||
1725 | This effect could also be achieved using barrier(), but READ_ONCE() | |
1726 | and WRITE_ONCE() are more selective: With READ_ONCE() and | |
1727 | WRITE_ONCE(), the compiler need only forget the contents of the | |
1728 | indicated memory locations, while with barrier() the compiler must | |
1729 | discard the value of all memory locations that it has currented | |
1730 | cached in any machine registers. Of course, the compiler must also | |
1731 | respect the order in which the READ_ONCE()s and WRITE_ONCE()s occur, | |
1732 | though the CPU of course need not do so. | |
692118da PM |
1733 | |
1734 | (*) The compiler is within its rights to invent stores to a variable, | |
1735 | as in the following example: | |
1736 | ||
1737 | if (a) | |
1738 | b = a; | |
1739 | else | |
1740 | b = 42; | |
1741 | ||
1742 | The compiler might save a branch by optimizing this as follows: | |
1743 | ||
1744 | b = 42; | |
1745 | if (a) | |
1746 | b = a; | |
1747 | ||
1748 | In single-threaded code, this is not only safe, but also saves | |
1749 | a branch. Unfortunately, in concurrent code, this optimization | |
1750 | could cause some other CPU to see a spurious value of 42 -- even | |
1751 | if variable 'a' was never zero -- when loading variable 'b'. | |
9af194ce | 1752 | Use WRITE_ONCE() to prevent this as follows: |
692118da PM |
1753 | |
1754 | if (a) | |
9af194ce | 1755 | WRITE_ONCE(b, a); |
692118da | 1756 | else |
9af194ce | 1757 | WRITE_ONCE(b, 42); |
692118da PM |
1758 | |
1759 | The compiler can also invent loads. These are usually less | |
1760 | damaging, but they can result in cache-line bouncing and thus in | |
9af194ce | 1761 | poor performance and scalability. Use READ_ONCE() to prevent |
692118da PM |
1762 | invented loads. |
1763 | ||
1764 | (*) For aligned memory locations whose size allows them to be accessed | |
1765 | with a single memory-reference instruction, prevents "load tearing" | |
1766 | and "store tearing," in which a single large access is replaced by | |
1767 | multiple smaller accesses. For example, given an architecture having | |
1768 | 16-bit store instructions with 7-bit immediate fields, the compiler | |
1769 | might be tempted to use two 16-bit store-immediate instructions to | |
1770 | implement the following 32-bit store: | |
1771 | ||
1772 | p = 0x00010002; | |
1773 | ||
1774 | Please note that GCC really does use this sort of optimization, | |
1775 | which is not surprising given that it would likely take more | |
1776 | than two instructions to build the constant and then store it. | |
1777 | This optimization can therefore be a win in single-threaded code. | |
1778 | In fact, a recent bug (since fixed) caused GCC to incorrectly use | |
1779 | this optimization in a volatile store. In the absence of such bugs, | |
9af194ce | 1780 | use of WRITE_ONCE() prevents store tearing in the following example: |
692118da | 1781 | |
9af194ce | 1782 | WRITE_ONCE(p, 0x00010002); |
692118da PM |
1783 | |
1784 | Use of packed structures can also result in load and store tearing, | |
1785 | as in this example: | |
1786 | ||
1787 | struct __attribute__((__packed__)) foo { | |
1788 | short a; | |
1789 | int b; | |
1790 | short c; | |
1791 | }; | |
1792 | struct foo foo1, foo2; | |
1793 | ... | |
1794 | ||
1795 | foo2.a = foo1.a; | |
1796 | foo2.b = foo1.b; | |
1797 | foo2.c = foo1.c; | |
1798 | ||
9af194ce PM |
1799 | Because there are no READ_ONCE() or WRITE_ONCE() wrappers and no |
1800 | volatile markings, the compiler would be well within its rights to | |
1801 | implement these three assignment statements as a pair of 32-bit | |
1802 | loads followed by a pair of 32-bit stores. This would result in | |
1803 | load tearing on 'foo1.b' and store tearing on 'foo2.b'. READ_ONCE() | |
1804 | and WRITE_ONCE() again prevent tearing in this example: | |
692118da PM |
1805 | |
1806 | foo2.a = foo1.a; | |
9af194ce | 1807 | WRITE_ONCE(foo2.b, READ_ONCE(foo1.b)); |
692118da PM |
1808 | foo2.c = foo1.c; |
1809 | ||
9af194ce PM |
1810 | All that aside, it is never necessary to use READ_ONCE() and |
1811 | WRITE_ONCE() on a variable that has been marked volatile. For example, | |
1812 | because 'jiffies' is marked volatile, it is never necessary to | |
1813 | say READ_ONCE(jiffies). The reason for this is that READ_ONCE() and | |
1814 | WRITE_ONCE() are implemented as volatile casts, which has no effect when | |
1815 | its argument is already marked volatile. | |
692118da PM |
1816 | |
1817 | Please note that these compiler barriers have no direct effect on the CPU, | |
1818 | which may then reorder things however it wishes. | |
108b42b4 DH |
1819 | |
1820 | ||
1821 | CPU MEMORY BARRIERS | |
1822 | ------------------- | |
1823 | ||
1824 | The Linux kernel has eight basic CPU memory barriers: | |
1825 | ||
1826 | TYPE MANDATORY SMP CONDITIONAL | |
1827 | =============== ======================= =========================== | |
1828 | GENERAL mb() smp_mb() | |
1829 | WRITE wmb() smp_wmb() | |
1830 | READ rmb() smp_rmb() | |
9ad3c143 | 1831 | DATA DEPENDENCY READ_ONCE() |
108b42b4 DH |
1832 | |
1833 | ||
73f10281 | 1834 | All memory barriers except the data dependency barriers imply a compiler |
0b6fa347 | 1835 | barrier. Data dependencies do not impose any additional compiler ordering. |
73f10281 | 1836 | |
9af194ce PM |
1837 | Aside: In the case of data dependencies, the compiler would be expected |
1838 | to issue the loads in the correct order (eg. `a[b]` would have to load | |
1839 | the value of b before loading a[b]), however there is no guarantee in | |
1840 | the C specification that the compiler may not speculate the value of b | |
1841 | (eg. is equal to 1) and load a before b (eg. tmp = a[1]; if (b != 1) | |
0b6fa347 SP |
1842 | tmp = a[b]; ). There is also the problem of a compiler reloading b after |
1843 | having loaded a[b], thus having a newer copy of b than a[b]. A consensus | |
9af194ce PM |
1844 | has not yet been reached about these problems, however the READ_ONCE() |
1845 | macro is a good place to start looking. | |
108b42b4 DH |
1846 | |
1847 | SMP memory barriers are reduced to compiler barriers on uniprocessor compiled | |
81fc6323 | 1848 | systems because it is assumed that a CPU will appear to be self-consistent, |
108b42b4 | 1849 | and will order overlapping accesses correctly with respect to itself. |
6a65d263 | 1850 | However, see the subsection on "Virtual Machine Guests" below. |
108b42b4 DH |
1851 | |
1852 | [!] Note that SMP memory barriers _must_ be used to control the ordering of | |
1853 | references to shared memory on SMP systems, though the use of locking instead | |
1854 | is sufficient. | |
1855 | ||
1856 | Mandatory barriers should not be used to control SMP effects, since mandatory | |
6a65d263 MT |
1857 | barriers impose unnecessary overhead on both SMP and UP systems. They may, |
1858 | however, be used to control MMIO effects on accesses through relaxed memory I/O | |
1859 | windows. These barriers are required even on non-SMP systems as they affect | |
1860 | the order in which memory operations appear to a device by prohibiting both the | |
1861 | compiler and the CPU from reordering them. | |
108b42b4 DH |
1862 | |
1863 | ||
1864 | There are some more advanced barrier functions: | |
1865 | ||
b92b8b35 | 1866 | (*) smp_store_mb(var, value) |
108b42b4 | 1867 | |
75b2bd55 | 1868 | This assigns the value to the variable and then inserts a full memory |
2d142e59 DB |
1869 | barrier after it. It isn't guaranteed to insert anything more than a |
1870 | compiler barrier in a UP compilation. | |
108b42b4 DH |
1871 | |
1872 | ||
1b15611e PZ |
1873 | (*) smp_mb__before_atomic(); |
1874 | (*) smp_mb__after_atomic(); | |
108b42b4 | 1875 | |
1b15611e PZ |
1876 | These are for use with atomic (such as add, subtract, increment and |
1877 | decrement) functions that don't return a value, especially when used for | |
1878 | reference counting. These functions do not imply memory barriers. | |
1879 | ||
1880 | These are also used for atomic bitop functions that do not return a | |
1881 | value (such as set_bit and clear_bit). | |
108b42b4 DH |
1882 | |
1883 | As an example, consider a piece of code that marks an object as being dead | |
1884 | and then decrements the object's reference count: | |
1885 | ||
1886 | obj->dead = 1; | |
1b15611e | 1887 | smp_mb__before_atomic(); |
108b42b4 DH |
1888 | atomic_dec(&obj->ref_count); |
1889 | ||
1890 | This makes sure that the death mark on the object is perceived to be set | |
1891 | *before* the reference counter is decremented. | |
1892 | ||
706eeb3e | 1893 | See Documentation/atomic_{t,bitops}.txt for more information. |
108b42b4 DH |
1894 | |
1895 | ||
1077fa36 AD |
1896 | (*) dma_wmb(); |
1897 | (*) dma_rmb(); | |
1898 | ||
1899 | These are for use with consistent memory to guarantee the ordering | |
1900 | of writes or reads of shared memory accessible to both the CPU and a | |
1901 | DMA capable device. | |
1902 | ||
1903 | For example, consider a device driver that shares memory with a device | |
1904 | and uses a descriptor status value to indicate if the descriptor belongs | |
1905 | to the device or the CPU, and a doorbell to notify it when new | |
1906 | descriptors are available: | |
1907 | ||
1908 | if (desc->status != DEVICE_OWN) { | |
1909 | /* do not read data until we own descriptor */ | |
1910 | dma_rmb(); | |
1911 | ||
1912 | /* read/modify data */ | |
1913 | read_data = desc->data; | |
1914 | desc->data = write_data; | |
1915 | ||
1916 | /* flush modifications before status update */ | |
1917 | dma_wmb(); | |
1918 | ||
1919 | /* assign ownership */ | |
1920 | desc->status = DEVICE_OWN; | |
1921 | ||
1077fa36 AD |
1922 | /* notify device of new descriptors */ |
1923 | writel(DESC_NOTIFY, doorbell); | |
1924 | } | |
1925 | ||
1926 | The dma_rmb() allows us guarantee the device has released ownership | |
7a458007 | 1927 | before we read the data from the descriptor, and the dma_wmb() allows |
1077fa36 | 1928 | us to guarantee the data is written to the descriptor before the device |
5846581e WD |
1929 | can see it now has ownership. Note that, when using writel(), a prior |
1930 | wmb() is not needed to guarantee that the cache coherent memory writes | |
1931 | have completed before writing to the MMIO region. The cheaper | |
1932 | writel_relaxed() does not provide this guarantee and must not be used | |
1933 | here. | |
1934 | ||
1935 | See the subsection "Kernel I/O barrier effects" for more information on | |
1936 | relaxed I/O accessors and the Documentation/DMA-API.txt file for more | |
1937 | information on consistent memory. | |
1077fa36 | 1938 | |
dfeccea6 | 1939 | |
108b42b4 DH |
1940 | MMIO WRITE BARRIER |
1941 | ------------------ | |
1942 | ||
1943 | The Linux kernel also has a special barrier for use with memory-mapped I/O | |
1944 | writes: | |
1945 | ||
1946 | mmiowb(); | |
1947 | ||
1948 | This is a variation on the mandatory write barrier that causes writes to weakly | |
1949 | ordered I/O regions to be partially ordered. Its effects may go beyond the | |
1950 | CPU->Hardware interface and actually affect the hardware at some level. | |
1951 | ||
166bda71 | 1952 | See the subsection "Acquires vs I/O accesses" for more information. |
108b42b4 DH |
1953 | |
1954 | ||
1955 | =============================== | |
1956 | IMPLICIT KERNEL MEMORY BARRIERS | |
1957 | =============================== | |
1958 | ||
1959 | Some of the other functions in the linux kernel imply memory barriers, amongst | |
670bd95e | 1960 | which are locking and scheduling functions. |
108b42b4 DH |
1961 | |
1962 | This specification is a _minimum_ guarantee; any particular architecture may | |
1963 | provide more substantial guarantees, but these may not be relied upon outside | |
1964 | of arch specific code. | |
1965 | ||
1966 | ||
166bda71 SP |
1967 | LOCK ACQUISITION FUNCTIONS |
1968 | -------------------------- | |
108b42b4 DH |
1969 | |
1970 | The Linux kernel has a number of locking constructs: | |
1971 | ||
1972 | (*) spin locks | |
1973 | (*) R/W spin locks | |
1974 | (*) mutexes | |
1975 | (*) semaphores | |
1976 | (*) R/W semaphores | |
108b42b4 | 1977 | |
2e4f5382 | 1978 | In all cases there are variants on "ACQUIRE" operations and "RELEASE" operations |
108b42b4 DH |
1979 | for each construct. These operations all imply certain barriers: |
1980 | ||
2e4f5382 | 1981 | (1) ACQUIRE operation implication: |
108b42b4 | 1982 | |
2e4f5382 PZ |
1983 | Memory operations issued after the ACQUIRE will be completed after the |
1984 | ACQUIRE operation has completed. | |
108b42b4 | 1985 | |
8dd853d7 | 1986 | Memory operations issued before the ACQUIRE may be completed after |
a9668cd6 | 1987 | the ACQUIRE operation has completed. |
108b42b4 | 1988 | |
2e4f5382 | 1989 | (2) RELEASE operation implication: |
108b42b4 | 1990 | |
2e4f5382 PZ |
1991 | Memory operations issued before the RELEASE will be completed before the |
1992 | RELEASE operation has completed. | |
108b42b4 | 1993 | |
2e4f5382 PZ |
1994 | Memory operations issued after the RELEASE may be completed before the |
1995 | RELEASE operation has completed. | |
108b42b4 | 1996 | |
2e4f5382 | 1997 | (3) ACQUIRE vs ACQUIRE implication: |
108b42b4 | 1998 | |
2e4f5382 PZ |
1999 | All ACQUIRE operations issued before another ACQUIRE operation will be |
2000 | completed before that ACQUIRE operation. | |
108b42b4 | 2001 | |
2e4f5382 | 2002 | (4) ACQUIRE vs RELEASE implication: |
108b42b4 | 2003 | |
2e4f5382 PZ |
2004 | All ACQUIRE operations issued before a RELEASE operation will be |
2005 | completed before the RELEASE operation. | |
108b42b4 | 2006 | |
2e4f5382 | 2007 | (5) Failed conditional ACQUIRE implication: |
108b42b4 | 2008 | |
2e4f5382 PZ |
2009 | Certain locking variants of the ACQUIRE operation may fail, either due to |
2010 | being unable to get the lock immediately, or due to receiving an unblocked | |
108b42b4 DH |
2011 | signal whilst asleep waiting for the lock to become available. Failed |
2012 | locks do not imply any sort of barrier. | |
2013 | ||
2e4f5382 PZ |
2014 | [!] Note: one of the consequences of lock ACQUIREs and RELEASEs being only |
2015 | one-way barriers is that the effects of instructions outside of a critical | |
2016 | section may seep into the inside of the critical section. | |
108b42b4 | 2017 | |
2e4f5382 PZ |
2018 | An ACQUIRE followed by a RELEASE may not be assumed to be full memory barrier |
2019 | because it is possible for an access preceding the ACQUIRE to happen after the | |
2020 | ACQUIRE, and an access following the RELEASE to happen before the RELEASE, and | |
2021 | the two accesses can themselves then cross: | |
670bd95e DH |
2022 | |
2023 | *A = a; | |
2e4f5382 PZ |
2024 | ACQUIRE M |
2025 | RELEASE M | |
670bd95e DH |
2026 | *B = b; |
2027 | ||
2028 | may occur as: | |
2029 | ||
2e4f5382 | 2030 | ACQUIRE M, STORE *B, STORE *A, RELEASE M |
17eb88e0 | 2031 | |
8dd853d7 PM |
2032 | When the ACQUIRE and RELEASE are a lock acquisition and release, |
2033 | respectively, this same reordering can occur if the lock's ACQUIRE and | |
2034 | RELEASE are to the same lock variable, but only from the perspective of | |
2035 | another CPU not holding that lock. In short, a ACQUIRE followed by an | |
2036 | RELEASE may -not- be assumed to be a full memory barrier. | |
2037 | ||
12d560f4 PM |
2038 | Similarly, the reverse case of a RELEASE followed by an ACQUIRE does |
2039 | not imply a full memory barrier. Therefore, the CPU's execution of the | |
2040 | critical sections corresponding to the RELEASE and the ACQUIRE can cross, | |
2041 | so that: | |
17eb88e0 PM |
2042 | |
2043 | *A = a; | |
2e4f5382 PZ |
2044 | RELEASE M |
2045 | ACQUIRE N | |
17eb88e0 PM |
2046 | *B = b; |
2047 | ||
2048 | could occur as: | |
2049 | ||
2e4f5382 | 2050 | ACQUIRE N, STORE *B, STORE *A, RELEASE M |
17eb88e0 | 2051 | |
8dd853d7 PM |
2052 | It might appear that this reordering could introduce a deadlock. |
2053 | However, this cannot happen because if such a deadlock threatened, | |
2054 | the RELEASE would simply complete, thereby avoiding the deadlock. | |
2055 | ||
2056 | Why does this work? | |
2057 | ||
2058 | One key point is that we are only talking about the CPU doing | |
2059 | the reordering, not the compiler. If the compiler (or, for | |
2060 | that matter, the developer) switched the operations, deadlock | |
2061 | -could- occur. | |
2062 | ||
2063 | But suppose the CPU reordered the operations. In this case, | |
2064 | the unlock precedes the lock in the assembly code. The CPU | |
2065 | simply elected to try executing the later lock operation first. | |
2066 | If there is a deadlock, this lock operation will simply spin (or | |
2067 | try to sleep, but more on that later). The CPU will eventually | |
2068 | execute the unlock operation (which preceded the lock operation | |
2069 | in the assembly code), which will unravel the potential deadlock, | |
2070 | allowing the lock operation to succeed. | |
2071 | ||
2072 | But what if the lock is a sleeplock? In that case, the code will | |
2073 | try to enter the scheduler, where it will eventually encounter | |
2074 | a memory barrier, which will force the earlier unlock operation | |
2075 | to complete, again unraveling the deadlock. There might be | |
2076 | a sleep-unlock race, but the locking primitive needs to resolve | |
2077 | such races properly in any case. | |
2078 | ||
108b42b4 DH |
2079 | Locks and semaphores may not provide any guarantee of ordering on UP compiled |
2080 | systems, and so cannot be counted on in such a situation to actually achieve | |
2081 | anything at all - especially with respect to I/O accesses - unless combined | |
2082 | with interrupt disabling operations. | |
2083 | ||
d7cab36d | 2084 | See also the section on "Inter-CPU acquiring barrier effects". |
108b42b4 DH |
2085 | |
2086 | ||
2087 | As an example, consider the following: | |
2088 | ||
2089 | *A = a; | |
2090 | *B = b; | |
2e4f5382 | 2091 | ACQUIRE |
108b42b4 DH |
2092 | *C = c; |
2093 | *D = d; | |
2e4f5382 | 2094 | RELEASE |
108b42b4 DH |
2095 | *E = e; |
2096 | *F = f; | |
2097 | ||
2098 | The following sequence of events is acceptable: | |
2099 | ||
2e4f5382 | 2100 | ACQUIRE, {*F,*A}, *E, {*C,*D}, *B, RELEASE |
108b42b4 DH |
2101 | |
2102 | [+] Note that {*F,*A} indicates a combined access. | |
2103 | ||
2104 | But none of the following are: | |
2105 | ||
2e4f5382 PZ |
2106 | {*F,*A}, *B, ACQUIRE, *C, *D, RELEASE, *E |
2107 | *A, *B, *C, ACQUIRE, *D, RELEASE, *E, *F | |
2108 | *A, *B, ACQUIRE, *C, RELEASE, *D, *E, *F | |
2109 | *B, ACQUIRE, *C, *D, RELEASE, {*F,*A}, *E | |
108b42b4 DH |
2110 | |
2111 | ||
2112 | ||
2113 | INTERRUPT DISABLING FUNCTIONS | |
2114 | ----------------------------- | |
2115 | ||
2e4f5382 PZ |
2116 | Functions that disable interrupts (ACQUIRE equivalent) and enable interrupts |
2117 | (RELEASE equivalent) will act as compiler barriers only. So if memory or I/O | |
108b42b4 DH |
2118 | barriers are required in such a situation, they must be provided from some |
2119 | other means. | |
2120 | ||
2121 | ||
50fa610a DH |
2122 | SLEEP AND WAKE-UP FUNCTIONS |
2123 | --------------------------- | |
2124 | ||
2125 | Sleeping and waking on an event flagged in global data can be viewed as an | |
2126 | interaction between two pieces of data: the task state of the task waiting for | |
2127 | the event and the global data used to indicate the event. To make sure that | |
2128 | these appear to happen in the right order, the primitives to begin the process | |
2129 | of going to sleep, and the primitives to initiate a wake up imply certain | |
2130 | barriers. | |
2131 | ||
2132 | Firstly, the sleeper normally follows something like this sequence of events: | |
2133 | ||
2134 | for (;;) { | |
2135 | set_current_state(TASK_UNINTERRUPTIBLE); | |
2136 | if (event_indicated) | |
2137 | break; | |
2138 | schedule(); | |
2139 | } | |
2140 | ||
2141 | A general memory barrier is interpolated automatically by set_current_state() | |
2142 | after it has altered the task state: | |
2143 | ||
2144 | CPU 1 | |
2145 | =============================== | |
2146 | set_current_state(); | |
b92b8b35 | 2147 | smp_store_mb(); |
50fa610a DH |
2148 | STORE current->state |
2149 | <general barrier> | |
2150 | LOAD event_indicated | |
2151 | ||
2152 | set_current_state() may be wrapped by: | |
2153 | ||
2154 | prepare_to_wait(); | |
2155 | prepare_to_wait_exclusive(); | |
2156 | ||
2157 | which therefore also imply a general memory barrier after setting the state. | |
2158 | The whole sequence above is available in various canned forms, all of which | |
2159 | interpolate the memory barrier in the right place: | |
2160 | ||
2161 | wait_event(); | |
2162 | wait_event_interruptible(); | |
2163 | wait_event_interruptible_exclusive(); | |
2164 | wait_event_interruptible_timeout(); | |
2165 | wait_event_killable(); | |
2166 | wait_event_timeout(); | |
2167 | wait_on_bit(); | |
2168 | wait_on_bit_lock(); | |
2169 | ||
2170 | ||
2171 | Secondly, code that performs a wake up normally follows something like this: | |
2172 | ||
2173 | event_indicated = 1; | |
2174 | wake_up(&event_wait_queue); | |
2175 | ||
2176 | or: | |
2177 | ||
2178 | event_indicated = 1; | |
2179 | wake_up_process(event_daemon); | |
2180 | ||
7696f991 AP |
2181 | A general memory barrier is executed by wake_up() if it wakes something up. |
2182 | If it doesn't wake anything up then a memory barrier may or may not be | |
2183 | executed; you must not rely on it. The barrier occurs before the task state | |
2184 | is accessed, in particular, it sits between the STORE to indicate the event | |
2185 | and the STORE to set TASK_RUNNING: | |
50fa610a | 2186 | |
7696f991 | 2187 | CPU 1 (Sleeper) CPU 2 (Waker) |
50fa610a DH |
2188 | =============================== =============================== |
2189 | set_current_state(); STORE event_indicated | |
b92b8b35 | 2190 | smp_store_mb(); wake_up(); |
7696f991 AP |
2191 | STORE current->state ... |
2192 | <general barrier> <general barrier> | |
2193 | LOAD event_indicated if ((LOAD task->state) & TASK_NORMAL) | |
2194 | STORE task->state | |
50fa610a | 2195 | |
7696f991 AP |
2196 | where "task" is the thread being woken up and it equals CPU 1's "current". |
2197 | ||
2198 | To repeat, a general memory barrier is guaranteed to be executed by wake_up() | |
2199 | if something is actually awakened, but otherwise there is no such guarantee. | |
2200 | To see this, consider the following sequence of events, where X and Y are both | |
2201 | initially zero: | |
5726ce06 PM |
2202 | |
2203 | CPU 1 CPU 2 | |
2204 | =============================== =============================== | |
7696f991 | 2205 | X = 1; Y = 1; |
5726ce06 | 2206 | smp_mb(); wake_up(); |
7696f991 AP |
2207 | LOAD Y LOAD X |
2208 | ||
2209 | If a wakeup does occur, one (at least) of the two loads must see 1. If, on | |
2210 | the other hand, a wakeup does not occur, both loads might see 0. | |
5726ce06 | 2211 | |
7696f991 AP |
2212 | wake_up_process() always executes a general memory barrier. The barrier again |
2213 | occurs before the task state is accessed. In particular, if the wake_up() in | |
2214 | the previous snippet were replaced by a call to wake_up_process() then one of | |
2215 | the two loads would be guaranteed to see 1. | |
5726ce06 | 2216 | |
50fa610a DH |
2217 | The available waker functions include: |
2218 | ||
2219 | complete(); | |
2220 | wake_up(); | |
2221 | wake_up_all(); | |
2222 | wake_up_bit(); | |
2223 | wake_up_interruptible(); | |
2224 | wake_up_interruptible_all(); | |
2225 | wake_up_interruptible_nr(); | |
2226 | wake_up_interruptible_poll(); | |
2227 | wake_up_interruptible_sync(); | |
2228 | wake_up_interruptible_sync_poll(); | |
2229 | wake_up_locked(); | |
2230 | wake_up_locked_poll(); | |
2231 | wake_up_nr(); | |
2232 | wake_up_poll(); | |
2233 | wake_up_process(); | |
2234 | ||
7696f991 AP |
2235 | In terms of memory ordering, these functions all provide the same guarantees of |
2236 | a wake_up() (or stronger). | |
50fa610a DH |
2237 | |
2238 | [!] Note that the memory barriers implied by the sleeper and the waker do _not_ | |
2239 | order multiple stores before the wake-up with respect to loads of those stored | |
2240 | values after the sleeper has called set_current_state(). For instance, if the | |
2241 | sleeper does: | |
2242 | ||
2243 | set_current_state(TASK_INTERRUPTIBLE); | |
2244 | if (event_indicated) | |
2245 | break; | |
2246 | __set_current_state(TASK_RUNNING); | |
2247 | do_something(my_data); | |
2248 | ||
2249 | and the waker does: | |
2250 | ||
2251 | my_data = value; | |
2252 | event_indicated = 1; | |
2253 | wake_up(&event_wait_queue); | |
2254 | ||
2255 | there's no guarantee that the change to event_indicated will be perceived by | |
2256 | the sleeper as coming after the change to my_data. In such a circumstance, the | |
2257 | code on both sides must interpolate its own memory barriers between the | |
2258 | separate data accesses. Thus the above sleeper ought to do: | |
2259 | ||
2260 | set_current_state(TASK_INTERRUPTIBLE); | |
2261 | if (event_indicated) { | |
2262 | smp_rmb(); | |
2263 | do_something(my_data); | |
2264 | } | |
2265 | ||
2266 | and the waker should do: | |
2267 | ||
2268 | my_data = value; | |
2269 | smp_wmb(); | |
2270 | event_indicated = 1; | |
2271 | wake_up(&event_wait_queue); | |
2272 | ||
2273 | ||
108b42b4 DH |
2274 | MISCELLANEOUS FUNCTIONS |
2275 | ----------------------- | |
2276 | ||
2277 | Other functions that imply barriers: | |
2278 | ||
2279 | (*) schedule() and similar imply full memory barriers. | |
2280 | ||
108b42b4 | 2281 | |
2e4f5382 PZ |
2282 | =================================== |
2283 | INTER-CPU ACQUIRING BARRIER EFFECTS | |
2284 | =================================== | |
108b42b4 DH |
2285 | |
2286 | On SMP systems locking primitives give a more substantial form of barrier: one | |
2287 | that does affect memory access ordering on other CPUs, within the context of | |
2288 | conflict on any particular lock. | |
2289 | ||
2290 | ||
2e4f5382 PZ |
2291 | ACQUIRES VS MEMORY ACCESSES |
2292 | --------------------------- | |
108b42b4 | 2293 | |
79afecfa | 2294 | Consider the following: the system has a pair of spinlocks (M) and (Q), and |
108b42b4 DH |
2295 | three CPUs; then should the following sequence of events occur: |
2296 | ||
2297 | CPU 1 CPU 2 | |
2298 | =============================== =============================== | |
9af194ce | 2299 | WRITE_ONCE(*A, a); WRITE_ONCE(*E, e); |
2e4f5382 | 2300 | ACQUIRE M ACQUIRE Q |
9af194ce PM |
2301 | WRITE_ONCE(*B, b); WRITE_ONCE(*F, f); |
2302 | WRITE_ONCE(*C, c); WRITE_ONCE(*G, g); | |
2e4f5382 | 2303 | RELEASE M RELEASE Q |
9af194ce | 2304 | WRITE_ONCE(*D, d); WRITE_ONCE(*H, h); |
108b42b4 | 2305 | |
81fc6323 | 2306 | Then there is no guarantee as to what order CPU 3 will see the accesses to *A |
108b42b4 | 2307 | through *H occur in, other than the constraints imposed by the separate locks |
0b6fa347 | 2308 | on the separate CPUs. It might, for example, see: |
108b42b4 | 2309 | |
2e4f5382 | 2310 | *E, ACQUIRE M, ACQUIRE Q, *G, *C, *F, *A, *B, RELEASE Q, *D, *H, RELEASE M |
108b42b4 DH |
2311 | |
2312 | But it won't see any of: | |
2313 | ||
2e4f5382 PZ |
2314 | *B, *C or *D preceding ACQUIRE M |
2315 | *A, *B or *C following RELEASE M | |
2316 | *F, *G or *H preceding ACQUIRE Q | |
2317 | *E, *F or *G following RELEASE Q | |
108b42b4 DH |
2318 | |
2319 | ||
108b42b4 | 2320 | |
2e4f5382 PZ |
2321 | ACQUIRES VS I/O ACCESSES |
2322 | ------------------------ | |
108b42b4 DH |
2323 | |
2324 | Under certain circumstances (especially involving NUMA), I/O accesses within | |
2325 | two spinlocked sections on two different CPUs may be seen as interleaved by the | |
2326 | PCI bridge, because the PCI bridge does not necessarily participate in the | |
2327 | cache-coherence protocol, and is therefore incapable of issuing the required | |
2328 | read memory barriers. | |
2329 | ||
2330 | For example: | |
2331 | ||
2332 | CPU 1 CPU 2 | |
2333 | =============================== =============================== | |
2334 | spin_lock(Q) | |
2335 | writel(0, ADDR) | |
2336 | writel(1, DATA); | |
2337 | spin_unlock(Q); | |
2338 | spin_lock(Q); | |
2339 | writel(4, ADDR); | |
2340 | writel(5, DATA); | |
2341 | spin_unlock(Q); | |
2342 | ||
2343 | may be seen by the PCI bridge as follows: | |
2344 | ||
2345 | STORE *ADDR = 0, STORE *ADDR = 4, STORE *DATA = 1, STORE *DATA = 5 | |
2346 | ||
2347 | which would probably cause the hardware to malfunction. | |
2348 | ||
2349 | ||
2350 | What is necessary here is to intervene with an mmiowb() before dropping the | |
2351 | spinlock, for example: | |
2352 | ||
2353 | CPU 1 CPU 2 | |
2354 | =============================== =============================== | |
2355 | spin_lock(Q) | |
2356 | writel(0, ADDR) | |
2357 | writel(1, DATA); | |
2358 | mmiowb(); | |
2359 | spin_unlock(Q); | |
2360 | spin_lock(Q); | |
2361 | writel(4, ADDR); | |
2362 | writel(5, DATA); | |
2363 | mmiowb(); | |
2364 | spin_unlock(Q); | |
2365 | ||
81fc6323 JP |
2366 | this will ensure that the two stores issued on CPU 1 appear at the PCI bridge |
2367 | before either of the stores issued on CPU 2. | |
108b42b4 DH |
2368 | |
2369 | ||
81fc6323 JP |
2370 | Furthermore, following a store by a load from the same device obviates the need |
2371 | for the mmiowb(), because the load forces the store to complete before the load | |
108b42b4 DH |
2372 | is performed: |
2373 | ||
2374 | CPU 1 CPU 2 | |
2375 | =============================== =============================== | |
2376 | spin_lock(Q) | |
2377 | writel(0, ADDR) | |
2378 | a = readl(DATA); | |
2379 | spin_unlock(Q); | |
2380 | spin_lock(Q); | |
2381 | writel(4, ADDR); | |
2382 | b = readl(DATA); | |
2383 | spin_unlock(Q); | |
2384 | ||
2385 | ||
0fe397f0 | 2386 | See Documentation/driver-api/device-io.rst for more information. |
108b42b4 DH |
2387 | |
2388 | ||
2389 | ================================= | |
2390 | WHERE ARE MEMORY BARRIERS NEEDED? | |
2391 | ================================= | |
2392 | ||
2393 | Under normal operation, memory operation reordering is generally not going to | |
2394 | be a problem as a single-threaded linear piece of code will still appear to | |
50fa610a | 2395 | work correctly, even if it's in an SMP kernel. There are, however, four |
108b42b4 DH |
2396 | circumstances in which reordering definitely _could_ be a problem: |
2397 | ||
2398 | (*) Interprocessor interaction. | |
2399 | ||
2400 | (*) Atomic operations. | |
2401 | ||
81fc6323 | 2402 | (*) Accessing devices. |
108b42b4 DH |
2403 | |
2404 | (*) Interrupts. | |
2405 | ||
2406 | ||
2407 | INTERPROCESSOR INTERACTION | |
2408 | -------------------------- | |
2409 | ||
2410 | When there's a system with more than one processor, more than one CPU in the | |
2411 | system may be working on the same data set at the same time. This can cause | |
2412 | synchronisation problems, and the usual way of dealing with them is to use | |
2413 | locks. Locks, however, are quite expensive, and so it may be preferable to | |
2414 | operate without the use of a lock if at all possible. In such a case | |
2415 | operations that affect both CPUs may have to be carefully ordered to prevent | |
2416 | a malfunction. | |
2417 | ||
2418 | Consider, for example, the R/W semaphore slow path. Here a waiting process is | |
2419 | queued on the semaphore, by virtue of it having a piece of its stack linked to | |
2420 | the semaphore's list of waiting processes: | |
2421 | ||
2422 | struct rw_semaphore { | |
2423 | ... | |
2424 | spinlock_t lock; | |
2425 | struct list_head waiters; | |
2426 | }; | |
2427 | ||
2428 | struct rwsem_waiter { | |
2429 | struct list_head list; | |
2430 | struct task_struct *task; | |
2431 | }; | |
2432 | ||
2433 | To wake up a particular waiter, the up_read() or up_write() functions have to: | |
2434 | ||
2435 | (1) read the next pointer from this waiter's record to know as to where the | |
2436 | next waiter record is; | |
2437 | ||
81fc6323 | 2438 | (2) read the pointer to the waiter's task structure; |
108b42b4 DH |
2439 | |
2440 | (3) clear the task pointer to tell the waiter it has been given the semaphore; | |
2441 | ||
2442 | (4) call wake_up_process() on the task; and | |
2443 | ||
2444 | (5) release the reference held on the waiter's task struct. | |
2445 | ||
81fc6323 | 2446 | In other words, it has to perform this sequence of events: |
108b42b4 DH |
2447 | |
2448 | LOAD waiter->list.next; | |
2449 | LOAD waiter->task; | |
2450 | STORE waiter->task; | |
2451 | CALL wakeup | |
2452 | RELEASE task | |
2453 | ||
2454 | and if any of these steps occur out of order, then the whole thing may | |
2455 | malfunction. | |
2456 | ||
2457 | Once it has queued itself and dropped the semaphore lock, the waiter does not | |
2458 | get the lock again; it instead just waits for its task pointer to be cleared | |
2459 | before proceeding. Since the record is on the waiter's stack, this means that | |
2460 | if the task pointer is cleared _before_ the next pointer in the list is read, | |
2461 | another CPU might start processing the waiter and might clobber the waiter's | |
2462 | stack before the up*() function has a chance to read the next pointer. | |
2463 | ||
2464 | Consider then what might happen to the above sequence of events: | |
2465 | ||
2466 | CPU 1 CPU 2 | |
2467 | =============================== =============================== | |
2468 | down_xxx() | |
2469 | Queue waiter | |
2470 | Sleep | |
2471 | up_yyy() | |
2472 | LOAD waiter->task; | |
2473 | STORE waiter->task; | |
2474 | Woken up by other event | |
2475 | <preempt> | |
2476 | Resume processing | |
2477 | down_xxx() returns | |
2478 | call foo() | |
2479 | foo() clobbers *waiter | |
2480 | </preempt> | |
2481 | LOAD waiter->list.next; | |
2482 | --- OOPS --- | |
2483 | ||
2484 | This could be dealt with using the semaphore lock, but then the down_xxx() | |
2485 | function has to needlessly get the spinlock again after being woken up. | |
2486 | ||
2487 | The way to deal with this is to insert a general SMP memory barrier: | |
2488 | ||
2489 | LOAD waiter->list.next; | |
2490 | LOAD waiter->task; | |
2491 | smp_mb(); | |
2492 | STORE waiter->task; | |
2493 | CALL wakeup | |
2494 | RELEASE task | |
2495 | ||
2496 | In this case, the barrier makes a guarantee that all memory accesses before the | |
2497 | barrier will appear to happen before all the memory accesses after the barrier | |
2498 | with respect to the other CPUs on the system. It does _not_ guarantee that all | |
2499 | the memory accesses before the barrier will be complete by the time the barrier | |
2500 | instruction itself is complete. | |
2501 | ||
2502 | On a UP system - where this wouldn't be a problem - the smp_mb() is just a | |
2503 | compiler barrier, thus making sure the compiler emits the instructions in the | |
6bc39274 DH |
2504 | right order without actually intervening in the CPU. Since there's only one |
2505 | CPU, that CPU's dependency ordering logic will take care of everything else. | |
108b42b4 DH |
2506 | |
2507 | ||
2508 | ATOMIC OPERATIONS | |
2509 | ----------------- | |
2510 | ||
dbc8700e DH |
2511 | Whilst they are technically interprocessor interaction considerations, atomic |
2512 | operations are noted specially as some of them imply full memory barriers and | |
2513 | some don't, but they're very heavily relied on as a group throughout the | |
2514 | kernel. | |
2515 | ||
706eeb3e | 2516 | See Documentation/atomic_t.txt for more information. |
108b42b4 DH |
2517 | |
2518 | ||
2519 | ACCESSING DEVICES | |
2520 | ----------------- | |
2521 | ||
2522 | Many devices can be memory mapped, and so appear to the CPU as if they're just | |
2523 | a set of memory locations. To control such a device, the driver usually has to | |
2524 | make the right memory accesses in exactly the right order. | |
2525 | ||
2526 | However, having a clever CPU or a clever compiler creates a potential problem | |
2527 | in that the carefully sequenced accesses in the driver code won't reach the | |
2528 | device in the requisite order if the CPU or the compiler thinks it is more | |
2529 | efficient to reorder, combine or merge accesses - something that would cause | |
2530 | the device to malfunction. | |
2531 | ||
2532 | Inside of the Linux kernel, I/O should be done through the appropriate accessor | |
2533 | routines - such as inb() or writel() - which know how to make such accesses | |
2534 | appropriately sequential. Whilst this, for the most part, renders the explicit | |
2535 | use of memory barriers unnecessary, there are a couple of situations where they | |
2536 | might be needed: | |
2537 | ||
2538 | (1) On some systems, I/O stores are not strongly ordered across all CPUs, and | |
2539 | so for _all_ general drivers locks should be used and mmiowb() must be | |
2540 | issued prior to unlocking the critical section. | |
2541 | ||
2542 | (2) If the accessor functions are used to refer to an I/O memory window with | |
2543 | relaxed memory access properties, then _mandatory_ memory barriers are | |
2544 | required to enforce ordering. | |
2545 | ||
0fe397f0 | 2546 | See Documentation/driver-api/device-io.rst for more information. |
108b42b4 DH |
2547 | |
2548 | ||
2549 | INTERRUPTS | |
2550 | ---------- | |
2551 | ||
2552 | A driver may be interrupted by its own interrupt service routine, and thus the | |
2553 | two parts of the driver may interfere with each other's attempts to control or | |
2554 | access the device. | |
2555 | ||
2556 | This may be alleviated - at least in part - by disabling local interrupts (a | |
2557 | form of locking), such that the critical operations are all contained within | |
2558 | the interrupt-disabled section in the driver. Whilst the driver's interrupt | |
2559 | routine is executing, the driver's core may not run on the same CPU, and its | |
2560 | interrupt is not permitted to happen again until the current interrupt has been | |
2561 | handled, thus the interrupt handler does not need to lock against that. | |
2562 | ||
2563 | However, consider a driver that was talking to an ethernet card that sports an | |
2564 | address register and a data register. If that driver's core talks to the card | |
2565 | under interrupt-disablement and then the driver's interrupt handler is invoked: | |
2566 | ||
2567 | LOCAL IRQ DISABLE | |
2568 | writew(ADDR, 3); | |
2569 | writew(DATA, y); | |
2570 | LOCAL IRQ ENABLE | |
2571 | <interrupt> | |
2572 | writew(ADDR, 4); | |
2573 | q = readw(DATA); | |
2574 | </interrupt> | |
2575 | ||
2576 | The store to the data register might happen after the second store to the | |
2577 | address register if ordering rules are sufficiently relaxed: | |
2578 | ||
2579 | STORE *ADDR = 3, STORE *ADDR = 4, STORE *DATA = y, q = LOAD *DATA | |
2580 | ||
2581 | ||
2582 | If ordering rules are relaxed, it must be assumed that accesses done inside an | |
2583 | interrupt disabled section may leak outside of it and may interleave with | |
2584 | accesses performed in an interrupt - and vice versa - unless implicit or | |
2585 | explicit barriers are used. | |
2586 | ||
2587 | Normally this won't be a problem because the I/O accesses done inside such | |
2588 | sections will include synchronous load operations on strictly ordered I/O | |
0b6fa347 | 2589 | registers that form implicit I/O barriers. If this isn't sufficient then an |
108b42b4 DH |
2590 | mmiowb() may need to be used explicitly. |
2591 | ||
2592 | ||
2593 | A similar situation may occur between an interrupt routine and two routines | |
0b6fa347 | 2594 | running on separate CPUs that communicate with each other. If such a case is |
108b42b4 DH |
2595 | likely, then interrupt-disabling locks should be used to guarantee ordering. |
2596 | ||
2597 | ||
2598 | ========================== | |
2599 | KERNEL I/O BARRIER EFFECTS | |
2600 | ========================== | |
2601 | ||
2602 | When accessing I/O memory, drivers should use the appropriate accessor | |
2603 | functions: | |
2604 | ||
2605 | (*) inX(), outX(): | |
2606 | ||
2607 | These are intended to talk to I/O space rather than memory space, but | |
0b6fa347 SP |
2608 | that's primarily a CPU-specific concept. The i386 and x86_64 processors |
2609 | do indeed have special I/O space access cycles and instructions, but many | |
108b42b4 DH |
2610 | CPUs don't have such a concept. |
2611 | ||
81fc6323 JP |
2612 | The PCI bus, amongst others, defines an I/O space concept which - on such |
2613 | CPUs as i386 and x86_64 - readily maps to the CPU's concept of I/O | |
6bc39274 DH |
2614 | space. However, it may also be mapped as a virtual I/O space in the CPU's |
2615 | memory map, particularly on those CPUs that don't support alternate I/O | |
2616 | spaces. | |
108b42b4 DH |
2617 | |
2618 | Accesses to this space may be fully synchronous (as on i386), but | |
2619 | intermediary bridges (such as the PCI host bridge) may not fully honour | |
2620 | that. | |
2621 | ||
2622 | They are guaranteed to be fully ordered with respect to each other. | |
2623 | ||
2624 | They are not guaranteed to be fully ordered with respect to other types of | |
2625 | memory and I/O operation. | |
2626 | ||
2627 | (*) readX(), writeX(): | |
2628 | ||
2629 | Whether these are guaranteed to be fully ordered and uncombined with | |
2630 | respect to each other on the issuing CPU depends on the characteristics | |
0b6fa347 | 2631 | defined for the memory window through which they're accessing. On later |
108b42b4 DH |
2632 | i386 architecture machines, for example, this is controlled by way of the |
2633 | MTRR registers. | |
2634 | ||
81fc6323 | 2635 | Ordinarily, these will be guaranteed to be fully ordered and uncombined, |
108b42b4 DH |
2636 | provided they're not accessing a prefetchable device. |
2637 | ||
2638 | However, intermediary hardware (such as a PCI bridge) may indulge in | |
2639 | deferral if it so wishes; to flush a store, a load from the same location | |
2640 | is preferred[*], but a load from the same device or from configuration | |
2641 | space should suffice for PCI. | |
2642 | ||
2643 | [*] NOTE! attempting to load from the same location as was written to may | |
e0edc78f IM |
2644 | cause a malfunction - consider the 16550 Rx/Tx serial registers for |
2645 | example. | |
108b42b4 DH |
2646 | |
2647 | Used with prefetchable I/O memory, an mmiowb() barrier may be required to | |
2648 | force stores to be ordered. | |
2649 | ||
2650 | Please refer to the PCI specification for more information on interactions | |
2651 | between PCI transactions. | |
2652 | ||
a8e0aead WD |
2653 | (*) readX_relaxed(), writeX_relaxed() |
2654 | ||
2655 | These are similar to readX() and writeX(), but provide weaker memory | |
0b6fa347 | 2656 | ordering guarantees. Specifically, they do not guarantee ordering with |
a8e0aead | 2657 | respect to normal memory accesses (e.g. DMA buffers) nor do they guarantee |
0b6fa347 SP |
2658 | ordering with respect to LOCK or UNLOCK operations. If the latter is |
2659 | required, an mmiowb() barrier can be used. Note that relaxed accesses to | |
a8e0aead WD |
2660 | the same peripheral are guaranteed to be ordered with respect to each |
2661 | other. | |
108b42b4 DH |
2662 | |
2663 | (*) ioreadX(), iowriteX() | |
2664 | ||
81fc6323 | 2665 | These will perform appropriately for the type of access they're actually |
108b42b4 DH |
2666 | doing, be it inX()/outX() or readX()/writeX(). |
2667 | ||
2668 | ||
2669 | ======================================== | |
2670 | ASSUMED MINIMUM EXECUTION ORDERING MODEL | |
2671 | ======================================== | |
2672 | ||
2673 | It has to be assumed that the conceptual CPU is weakly-ordered but that it will | |
2674 | maintain the appearance of program causality with respect to itself. Some CPUs | |
2675 | (such as i386 or x86_64) are more constrained than others (such as powerpc or | |
2676 | frv), and so the most relaxed case (namely DEC Alpha) must be assumed outside | |
2677 | of arch-specific code. | |
2678 | ||
2679 | This means that it must be considered that the CPU will execute its instruction | |
2680 | stream in any order it feels like - or even in parallel - provided that if an | |
81fc6323 | 2681 | instruction in the stream depends on an earlier instruction, then that |
108b42b4 DH |
2682 | earlier instruction must be sufficiently complete[*] before the later |
2683 | instruction may proceed; in other words: provided that the appearance of | |
2684 | causality is maintained. | |
2685 | ||
2686 | [*] Some instructions have more than one effect - such as changing the | |
2687 | condition codes, changing registers or changing memory - and different | |
2688 | instructions may depend on different effects. | |
2689 | ||
2690 | A CPU may also discard any instruction sequence that winds up having no | |
2691 | ultimate effect. For example, if two adjacent instructions both load an | |
2692 | immediate value into the same register, the first may be discarded. | |
2693 | ||
2694 | ||
2695 | Similarly, it has to be assumed that compiler might reorder the instruction | |
2696 | stream in any way it sees fit, again provided the appearance of causality is | |
2697 | maintained. | |
2698 | ||
2699 | ||
2700 | ============================ | |
2701 | THE EFFECTS OF THE CPU CACHE | |
2702 | ============================ | |
2703 | ||
2704 | The way cached memory operations are perceived across the system is affected to | |
2705 | a certain extent by the caches that lie between CPUs and memory, and by the | |
2706 | memory coherence system that maintains the consistency of state in the system. | |
2707 | ||
2708 | As far as the way a CPU interacts with another part of the system through the | |
2709 | caches goes, the memory system has to include the CPU's caches, and memory | |
2710 | barriers for the most part act at the interface between the CPU and its cache | |
2711 | (memory barriers logically act on the dotted line in the following diagram): | |
2712 | ||
2713 | <--- CPU ---> : <----------- Memory -----------> | |
2714 | : | |
2715 | +--------+ +--------+ : +--------+ +-----------+ | |
2716 | | | | | : | | | | +--------+ | |
e0edc78f IM |
2717 | | CPU | | Memory | : | CPU | | | | | |
2718 | | Core |--->| Access |----->| Cache |<-->| | | | | |
108b42b4 | 2719 | | | | Queue | : | | | |--->| Memory | |
e0edc78f IM |
2720 | | | | | : | | | | | | |
2721 | +--------+ +--------+ : +--------+ | | | | | |
108b42b4 DH |
2722 | : | Cache | +--------+ |
2723 | : | Coherency | | |
2724 | : | Mechanism | +--------+ | |
2725 | +--------+ +--------+ : +--------+ | | | | | |
2726 | | | | | : | | | | | | | |
2727 | | CPU | | Memory | : | CPU | | |--->| Device | | |
e0edc78f IM |
2728 | | Core |--->| Access |----->| Cache |<-->| | | | |
2729 | | | | Queue | : | | | | | | | |
108b42b4 DH |
2730 | | | | | : | | | | +--------+ |
2731 | +--------+ +--------+ : +--------+ +-----------+ | |
2732 | : | |
2733 | : | |
2734 | ||
2735 | Although any particular load or store may not actually appear outside of the | |
2736 | CPU that issued it since it may have been satisfied within the CPU's own cache, | |
2737 | it will still appear as if the full memory access had taken place as far as the | |
2738 | other CPUs are concerned since the cache coherency mechanisms will migrate the | |
2739 | cacheline over to the accessing CPU and propagate the effects upon conflict. | |
2740 | ||
2741 | The CPU core may execute instructions in any order it deems fit, provided the | |
2742 | expected program causality appears to be maintained. Some of the instructions | |
2743 | generate load and store operations which then go into the queue of memory | |
2744 | accesses to be performed. The core may place these in the queue in any order | |
2745 | it wishes, and continue execution until it is forced to wait for an instruction | |
2746 | to complete. | |
2747 | ||
2748 | What memory barriers are concerned with is controlling the order in which | |
2749 | accesses cross from the CPU side of things to the memory side of things, and | |
2750 | the order in which the effects are perceived to happen by the other observers | |
2751 | in the system. | |
2752 | ||
2753 | [!] Memory barriers are _not_ needed within a given CPU, as CPUs always see | |
2754 | their own loads and stores as if they had happened in program order. | |
2755 | ||
2756 | [!] MMIO or other device accesses may bypass the cache system. This depends on | |
2757 | the properties of the memory window through which devices are accessed and/or | |
2758 | the use of any special device communication instructions the CPU may have. | |
2759 | ||
2760 | ||
2761 | CACHE COHERENCY | |
2762 | --------------- | |
2763 | ||
2764 | Life isn't quite as simple as it may appear above, however: for while the | |
2765 | caches are expected to be coherent, there's no guarantee that that coherency | |
2766 | will be ordered. This means that whilst changes made on one CPU will | |
2767 | eventually become visible on all CPUs, there's no guarantee that they will | |
2768 | become apparent in the same order on those other CPUs. | |
2769 | ||
2770 | ||
81fc6323 JP |
2771 | Consider dealing with a system that has a pair of CPUs (1 & 2), each of which |
2772 | has a pair of parallel data caches (CPU 1 has A/B, and CPU 2 has C/D): | |
108b42b4 DH |
2773 | |
2774 | : | |
2775 | : +--------+ | |
2776 | : +---------+ | | | |
2777 | +--------+ : +--->| Cache A |<------->| | | |
2778 | | | : | +---------+ | | | |
2779 | | CPU 1 |<---+ | | | |
2780 | | | : | +---------+ | | | |
2781 | +--------+ : +--->| Cache B |<------->| | | |
2782 | : +---------+ | | | |
2783 | : | Memory | | |
2784 | : +---------+ | System | | |
2785 | +--------+ : +--->| Cache C |<------->| | | |
2786 | | | : | +---------+ | | | |
2787 | | CPU 2 |<---+ | | | |
2788 | | | : | +---------+ | | | |
2789 | +--------+ : +--->| Cache D |<------->| | | |
2790 | : +---------+ | | | |
2791 | : +--------+ | |
2792 | : | |
2793 | ||
2794 | Imagine the system has the following properties: | |
2795 | ||
2796 | (*) an odd-numbered cache line may be in cache A, cache C or it may still be | |
2797 | resident in memory; | |
2798 | ||
2799 | (*) an even-numbered cache line may be in cache B, cache D or it may still be | |
2800 | resident in memory; | |
2801 | ||
2802 | (*) whilst the CPU core is interrogating one cache, the other cache may be | |
2803 | making use of the bus to access the rest of the system - perhaps to | |
2804 | displace a dirty cacheline or to do a speculative load; | |
2805 | ||
2806 | (*) each cache has a queue of operations that need to be applied to that cache | |
2807 | to maintain coherency with the rest of the system; | |
2808 | ||
2809 | (*) the coherency queue is not flushed by normal loads to lines already | |
2810 | present in the cache, even though the contents of the queue may | |
81fc6323 | 2811 | potentially affect those loads. |
108b42b4 DH |
2812 | |
2813 | Imagine, then, that two writes are made on the first CPU, with a write barrier | |
2814 | between them to guarantee that they will appear to reach that CPU's caches in | |
2815 | the requisite order: | |
2816 | ||
2817 | CPU 1 CPU 2 COMMENT | |
2818 | =============== =============== ======================================= | |
2819 | u == 0, v == 1 and p == &u, q == &u | |
2820 | v = 2; | |
81fc6323 | 2821 | smp_wmb(); Make sure change to v is visible before |
108b42b4 DH |
2822 | change to p |
2823 | <A:modify v=2> v is now in cache A exclusively | |
2824 | p = &v; | |
2825 | <B:modify p=&v> p is now in cache B exclusively | |
2826 | ||
2827 | The write memory barrier forces the other CPUs in the system to perceive that | |
2828 | the local CPU's caches have apparently been updated in the correct order. But | |
81fc6323 | 2829 | now imagine that the second CPU wants to read those values: |
108b42b4 DH |
2830 | |
2831 | CPU 1 CPU 2 COMMENT | |
2832 | =============== =============== ======================================= | |
2833 | ... | |
2834 | q = p; | |
2835 | x = *q; | |
2836 | ||
81fc6323 | 2837 | The above pair of reads may then fail to happen in the expected order, as the |
108b42b4 DH |
2838 | cacheline holding p may get updated in one of the second CPU's caches whilst |
2839 | the update to the cacheline holding v is delayed in the other of the second | |
2840 | CPU's caches by some other cache event: | |
2841 | ||
2842 | CPU 1 CPU 2 COMMENT | |
2843 | =============== =============== ======================================= | |
2844 | u == 0, v == 1 and p == &u, q == &u | |
2845 | v = 2; | |
2846 | smp_wmb(); | |
2847 | <A:modify v=2> <C:busy> | |
2848 | <C:queue v=2> | |
79afecfa | 2849 | p = &v; q = p; |
108b42b4 DH |
2850 | <D:request p> |
2851 | <B:modify p=&v> <D:commit p=&v> | |
e0edc78f | 2852 | <D:read p> |
108b42b4 DH |
2853 | x = *q; |
2854 | <C:read *q> Reads from v before v updated in cache | |
2855 | <C:unbusy> | |
2856 | <C:commit v=2> | |
2857 | ||
2858 | Basically, whilst both cachelines will be updated on CPU 2 eventually, there's | |
2859 | no guarantee that, without intervention, the order of update will be the same | |
2860 | as that committed on CPU 1. | |
2861 | ||
2862 | ||
2863 | To intervene, we need to interpolate a data dependency barrier or a read | |
f28f0868 PM |
2864 | barrier between the loads (which as of v4.15 is supplied unconditionally |
2865 | by the READ_ONCE() macro). This will force the cache to commit its | |
2866 | coherency queue before processing any further requests: | |
108b42b4 DH |
2867 | |
2868 | CPU 1 CPU 2 COMMENT | |
2869 | =============== =============== ======================================= | |
2870 | u == 0, v == 1 and p == &u, q == &u | |
2871 | v = 2; | |
2872 | smp_wmb(); | |
2873 | <A:modify v=2> <C:busy> | |
2874 | <C:queue v=2> | |
3fda982c | 2875 | p = &v; q = p; |
108b42b4 DH |
2876 | <D:request p> |
2877 | <B:modify p=&v> <D:commit p=&v> | |
e0edc78f | 2878 | <D:read p> |
108b42b4 DH |
2879 | smp_read_barrier_depends() |
2880 | <C:unbusy> | |
2881 | <C:commit v=2> | |
2882 | x = *q; | |
2883 | <C:read *q> Reads from v after v updated in cache | |
2884 | ||
2885 | ||
2886 | This sort of problem can be encountered on DEC Alpha processors as they have a | |
2887 | split cache that improves performance by making better use of the data bus. | |
2888 | Whilst most CPUs do imply a data dependency barrier on the read when a memory | |
2889 | access depends on a read, not all do, so it may not be relied on. | |
2890 | ||
2891 | Other CPUs may also have split caches, but must coordinate between the various | |
3f6dee9b | 2892 | cachelets for normal memory accesses. The semantics of the Alpha removes the |
9ad3c143 PM |
2893 | need for hardware coordination in the absence of memory barriers, which |
2894 | permitted Alpha to sport higher CPU clock rates back in the day. However, | |
f28f0868 PM |
2895 | please note that (again, as of v4.15) smp_read_barrier_depends() should not |
2896 | be used except in Alpha arch-specific code and within the READ_ONCE() macro. | |
108b42b4 DH |
2897 | |
2898 | ||
2899 | CACHE COHERENCY VS DMA | |
2900 | ---------------------- | |
2901 | ||
2902 | Not all systems maintain cache coherency with respect to devices doing DMA. In | |
2903 | such cases, a device attempting DMA may obtain stale data from RAM because | |
2904 | dirty cache lines may be resident in the caches of various CPUs, and may not | |
2905 | have been written back to RAM yet. To deal with this, the appropriate part of | |
2906 | the kernel must flush the overlapping bits of cache on each CPU (and maybe | |
2907 | invalidate them as well). | |
2908 | ||
2909 | In addition, the data DMA'd to RAM by a device may be overwritten by dirty | |
2910 | cache lines being written back to RAM from a CPU's cache after the device has | |
81fc6323 JP |
2911 | installed its own data, or cache lines present in the CPU's cache may simply |
2912 | obscure the fact that RAM has been updated, until at such time as the cacheline | |
2913 | is discarded from the CPU's cache and reloaded. To deal with this, the | |
2914 | appropriate part of the kernel must invalidate the overlapping bits of the | |
108b42b4 DH |
2915 | cache on each CPU. |
2916 | ||
de0f51e4 | 2917 | See Documentation/core-api/cachetlb.rst for more information on cache management. |
108b42b4 DH |
2918 | |
2919 | ||
2920 | CACHE COHERENCY VS MMIO | |
2921 | ----------------------- | |
2922 | ||
2923 | Memory mapped I/O usually takes place through memory locations that are part of | |
81fc6323 | 2924 | a window in the CPU's memory space that has different properties assigned than |
108b42b4 DH |
2925 | the usual RAM directed window. |
2926 | ||
2927 | Amongst these properties is usually the fact that such accesses bypass the | |
2928 | caching entirely and go directly to the device buses. This means MMIO accesses | |
2929 | may, in effect, overtake accesses to cached memory that were emitted earlier. | |
2930 | A memory barrier isn't sufficient in such a case, but rather the cache must be | |
2931 | flushed between the cached memory write and the MMIO access if the two are in | |
2932 | any way dependent. | |
2933 | ||
2934 | ||
2935 | ========================= | |
2936 | THE THINGS CPUS GET UP TO | |
2937 | ========================= | |
2938 | ||
2939 | A programmer might take it for granted that the CPU will perform memory | |
81fc6323 | 2940 | operations in exactly the order specified, so that if the CPU is, for example, |
108b42b4 DH |
2941 | given the following piece of code to execute: |
2942 | ||
9af194ce PM |
2943 | a = READ_ONCE(*A); |
2944 | WRITE_ONCE(*B, b); | |
2945 | c = READ_ONCE(*C); | |
2946 | d = READ_ONCE(*D); | |
2947 | WRITE_ONCE(*E, e); | |
108b42b4 | 2948 | |
81fc6323 | 2949 | they would then expect that the CPU will complete the memory operation for each |
108b42b4 DH |
2950 | instruction before moving on to the next one, leading to a definite sequence of |
2951 | operations as seen by external observers in the system: | |
2952 | ||
2953 | LOAD *A, STORE *B, LOAD *C, LOAD *D, STORE *E. | |
2954 | ||
2955 | ||
2956 | Reality is, of course, much messier. With many CPUs and compilers, the above | |
2957 | assumption doesn't hold because: | |
2958 | ||
2959 | (*) loads are more likely to need to be completed immediately to permit | |
2960 | execution progress, whereas stores can often be deferred without a | |
2961 | problem; | |
2962 | ||
2963 | (*) loads may be done speculatively, and the result discarded should it prove | |
2964 | to have been unnecessary; | |
2965 | ||
81fc6323 JP |
2966 | (*) loads may be done speculatively, leading to the result having been fetched |
2967 | at the wrong time in the expected sequence of events; | |
108b42b4 DH |
2968 | |
2969 | (*) the order of the memory accesses may be rearranged to promote better use | |
2970 | of the CPU buses and caches; | |
2971 | ||
2972 | (*) loads and stores may be combined to improve performance when talking to | |
2973 | memory or I/O hardware that can do batched accesses of adjacent locations, | |
2974 | thus cutting down on transaction setup costs (memory and PCI devices may | |
2975 | both be able to do this); and | |
2976 | ||
2977 | (*) the CPU's data cache may affect the ordering, and whilst cache-coherency | |
2978 | mechanisms may alleviate this - once the store has actually hit the cache | |
2979 | - there's no guarantee that the coherency management will be propagated in | |
2980 | order to other CPUs. | |
2981 | ||
2982 | So what another CPU, say, might actually observe from the above piece of code | |
2983 | is: | |
2984 | ||
2985 | LOAD *A, ..., LOAD {*C,*D}, STORE *E, STORE *B | |
2986 | ||
2987 | (Where "LOAD {*C,*D}" is a combined load) | |
2988 | ||
2989 | ||
2990 | However, it is guaranteed that a CPU will be self-consistent: it will see its | |
2991 | _own_ accesses appear to be correctly ordered, without the need for a memory | |
2992 | barrier. For instance with the following code: | |
2993 | ||
9af194ce PM |
2994 | U = READ_ONCE(*A); |
2995 | WRITE_ONCE(*A, V); | |
2996 | WRITE_ONCE(*A, W); | |
2997 | X = READ_ONCE(*A); | |
2998 | WRITE_ONCE(*A, Y); | |
2999 | Z = READ_ONCE(*A); | |
108b42b4 DH |
3000 | |
3001 | and assuming no intervention by an external influence, it can be assumed that | |
3002 | the final result will appear to be: | |
3003 | ||
3004 | U == the original value of *A | |
3005 | X == W | |
3006 | Z == Y | |
3007 | *A == Y | |
3008 | ||
3009 | The code above may cause the CPU to generate the full sequence of memory | |
3010 | accesses: | |
3011 | ||
3012 | U=LOAD *A, STORE *A=V, STORE *A=W, X=LOAD *A, STORE *A=Y, Z=LOAD *A | |
3013 | ||
3014 | in that order, but, without intervention, the sequence may have almost any | |
9af194ce PM |
3015 | combination of elements combined or discarded, provided the program's view |
3016 | of the world remains consistent. Note that READ_ONCE() and WRITE_ONCE() | |
3017 | are -not- optional in the above example, as there are architectures | |
3018 | where a given CPU might reorder successive loads to the same location. | |
3019 | On such architectures, READ_ONCE() and WRITE_ONCE() do whatever is | |
3020 | necessary to prevent this, for example, on Itanium the volatile casts | |
3021 | used by READ_ONCE() and WRITE_ONCE() cause GCC to emit the special ld.acq | |
3022 | and st.rel instructions (respectively) that prevent such reordering. | |
108b42b4 DH |
3023 | |
3024 | The compiler may also combine, discard or defer elements of the sequence before | |
3025 | the CPU even sees them. | |
3026 | ||
3027 | For instance: | |
3028 | ||
3029 | *A = V; | |
3030 | *A = W; | |
3031 | ||
3032 | may be reduced to: | |
3033 | ||
3034 | *A = W; | |
3035 | ||
9af194ce | 3036 | since, without either a write barrier or an WRITE_ONCE(), it can be |
2ecf8101 | 3037 | assumed that the effect of the storage of V to *A is lost. Similarly: |
108b42b4 DH |
3038 | |
3039 | *A = Y; | |
3040 | Z = *A; | |
3041 | ||
9af194ce PM |
3042 | may, without a memory barrier or an READ_ONCE() and WRITE_ONCE(), be |
3043 | reduced to: | |
108b42b4 DH |
3044 | |
3045 | *A = Y; | |
3046 | Z = Y; | |
3047 | ||
3048 | and the LOAD operation never appear outside of the CPU. | |
3049 | ||
3050 | ||
3051 | AND THEN THERE'S THE ALPHA | |
3052 | -------------------------- | |
3053 | ||
3054 | The DEC Alpha CPU is one of the most relaxed CPUs there is. Not only that, | |
3055 | some versions of the Alpha CPU have a split data cache, permitting them to have | |
81fc6323 | 3056 | two semantically-related cache lines updated at separate times. This is where |
108b42b4 DH |
3057 | the data dependency barrier really becomes necessary as this synchronises both |
3058 | caches with the memory coherence system, thus making it seem like pointer | |
3059 | changes vs new data occur in the right order. | |
3060 | ||
f28f0868 PM |
3061 | The Alpha defines the Linux kernel's memory model, although as of v4.15 |
3062 | the Linux kernel's addition of smp_read_barrier_depends() to READ_ONCE() | |
3063 | greatly reduced Alpha's impact on the memory model. | |
108b42b4 DH |
3064 | |
3065 | See the subsection on "Cache Coherency" above. | |
3066 | ||
0b6fa347 | 3067 | |
6a65d263 | 3068 | VIRTUAL MACHINE GUESTS |
3dbf0913 | 3069 | ---------------------- |
6a65d263 MT |
3070 | |
3071 | Guests running within virtual machines might be affected by SMP effects even if | |
3072 | the guest itself is compiled without SMP support. This is an artifact of | |
3073 | interfacing with an SMP host while running an UP kernel. Using mandatory | |
3074 | barriers for this use-case would be possible but is often suboptimal. | |
3075 | ||
3076 | To handle this case optimally, low-level virt_mb() etc macros are available. | |
3077 | These have the same effect as smp_mb() etc when SMP is enabled, but generate | |
0b6fa347 | 3078 | identical code for SMP and non-SMP systems. For example, virtual machine guests |
6a65d263 MT |
3079 | should use virt_mb() rather than smp_mb() when synchronizing against a |
3080 | (possibly SMP) host. | |
3081 | ||
3082 | These are equivalent to smp_mb() etc counterparts in all other respects, | |
3083 | in particular, they do not control MMIO effects: to control | |
3084 | MMIO effects, use mandatory barriers. | |
108b42b4 | 3085 | |
0b6fa347 | 3086 | |
90fddabf DH |
3087 | ============ |
3088 | EXAMPLE USES | |
3089 | ============ | |
3090 | ||
3091 | CIRCULAR BUFFERS | |
3092 | ---------------- | |
3093 | ||
3094 | Memory barriers can be used to implement circular buffering without the need | |
3095 | of a lock to serialise the producer with the consumer. See: | |
3096 | ||
d8a121e3 | 3097 | Documentation/core-api/circular-buffers.rst |
90fddabf DH |
3098 | |
3099 | for details. | |
3100 | ||
3101 | ||
108b42b4 DH |
3102 | ========== |
3103 | REFERENCES | |
3104 | ========== | |
3105 | ||
3106 | Alpha AXP Architecture Reference Manual, Second Edition (Sites & Witek, | |
3107 | Digital Press) | |
3108 | Chapter 5.2: Physical Address Space Characteristics | |
3109 | Chapter 5.4: Caches and Write Buffers | |
3110 | Chapter 5.5: Data Sharing | |
3111 | Chapter 5.6: Read/Write Ordering | |
3112 | ||
3113 | AMD64 Architecture Programmer's Manual Volume 2: System Programming | |
3114 | Chapter 7.1: Memory-Access Ordering | |
3115 | Chapter 7.4: Buffering and Combining Memory Writes | |
3116 | ||
f1ab25a3 PM |
3117 | ARM Architecture Reference Manual (ARMv8, for ARMv8-A architecture profile) |
3118 | Chapter B2: The AArch64 Application Level Memory Model | |
3119 | ||
108b42b4 DH |
3120 | IA-32 Intel Architecture Software Developer's Manual, Volume 3: |
3121 | System Programming Guide | |
3122 | Chapter 7.1: Locked Atomic Operations | |
3123 | Chapter 7.2: Memory Ordering | |
3124 | Chapter 7.4: Serializing Instructions | |
3125 | ||
3126 | The SPARC Architecture Manual, Version 9 | |
3127 | Chapter 8: Memory Models | |
3128 | Appendix D: Formal Specification of the Memory Models | |
3129 | Appendix J: Programming with the Memory Models | |
3130 | ||
f1ab25a3 PM |
3131 | Storage in the PowerPC (Stone and Fitzgerald) |
3132 | ||
108b42b4 DH |
3133 | UltraSPARC Programmer Reference Manual |
3134 | Chapter 5: Memory Accesses and Cacheability | |
3135 | Chapter 15: Sparc-V9 Memory Models | |
3136 | ||
3137 | UltraSPARC III Cu User's Manual | |
3138 | Chapter 9: Memory Models | |
3139 | ||
3140 | UltraSPARC IIIi Processor User's Manual | |
3141 | Chapter 8: Memory Models | |
3142 | ||
3143 | UltraSPARC Architecture 2005 | |
3144 | Chapter 9: Memory | |
3145 | Appendix D: Formal Specifications of the Memory Models | |
3146 | ||
3147 | UltraSPARC T1 Supplement to the UltraSPARC Architecture 2005 | |
3148 | Chapter 8: Memory Models | |
3149 | Appendix F: Caches and Cache Coherency | |
3150 | ||
3151 | Solaris Internals, Core Kernel Architecture, p63-68: | |
3152 | Chapter 3.3: Hardware Considerations for Locks and | |
3153 | Synchronization | |
3154 | ||
3155 | Unix Systems for Modern Architectures, Symmetric Multiprocessing and Caching | |
3156 | for Kernel Programmers: | |
3157 | Chapter 13: Other Memory Models | |
3158 | ||
3159 | Intel Itanium Architecture Software Developer's Manual: Volume 1: | |
3160 | Section 2.6: Speculation | |
3161 | Section 4.4: Memory Access |