1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
|
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
<html>
<head>
<title>LLVM Atomic Instructions and Concurrency Guide</title>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<link rel="stylesheet" href="llvm.css" type="text/css">
</head>
<body>
<h1>
LLVM Atomic Instructions and Concurrency Guide
</h1>
<ol>
<li><a href="#introduction">Introduction</a></li>
<li><a href="#outsideatomic">Optimization outside atomic</a></li>
<li><a href="#atomicinst">Atomic instructions</a></li>
<li><a href="#ordering">Atomic orderings</a></li>
<li><a href="#iropt">Atomics and IR optimization</a></li>
<li><a href="#codegen">Atomics and Codegen</a></li>
</ol>
<div class="doc_author">
<p>Written by Eli Friedman</p>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="introduction">Introduction</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p>Historically, LLVM has not had very strong support for concurrency; some
minimal intrinsics were provided, and <code>volatile</code> was used in some
cases to achieve rough semantics in the presence of concurrency. However, this
is changing; there are now new instructions which are well-defined in the
presence of threads and asynchronous signals, and the model for existing
instructions has been clarified in the IR.</p>
<p>The atomic instructions are designed specifically to provide readable IR and
optimized code generation for the following:</p>
<ul>
<li>The new C++0x <code><atomic></code> header.
(<a href="http://www.open-std.org/jtc1/sc22/wg21/">C++0x draft available here</a>.)
(<a href="http://www.open-std.org/jtc1/sc22/wg14/">C1x draft available here</a>)</li>
<li>Proper semantics for Java-style memory, for both <code>volatile</code> and
regular shared variables.
(<a href="http://java.sun.com/docs/books/jls/third_edition/html/memory.html">Java Specification</a>)</li>
<li>gcc-compatible <code>__sync_*</code> builtins.
(<a href="http://gcc.gnu.org/onlinedocs/gcc/Atomic-Builtins.html">Description</a>)</li>
<li>Other scenarios with atomic semantics, including <code>static</code>
variables with non-trivial constructors in C++.</li>
</ul>
<p>Atomic and volatile in the IR are orthogonal; "volatile" is the C/C++
volatile, which ensures that every volatile load and store happens and is
performed in the stated order. A couple examples: if a
SequentiallyConsistent store is immediately followed by another
SequentiallyConsistent store to the same address, the first store can
be erased. This transformation is not allowed for a pair of volatile
stores. On the other hand, a non-volatile non-atomic load can be moved
across a volatile load freely, but not an Acquire load.</p>
<p>This document is intended to provide a guide to anyone either writing a
frontend for LLVM or working on optimization passes for LLVM with a guide
for how to deal with instructions with special semantics in the presence of
concurrency. This is not intended to be a precise guide to the semantics;
the details can get extremely complicated and unreadable, and are not
usually necessary.</p>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="outsideatomic">Optimization outside atomic</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p>The basic <code>'load'</code> and <code>'store'</code> allow a variety of
optimizations, but can lead to undefined results in a concurrent environment;
see <a href="#o_nonatomic">NonAtomic</a>. This section specifically goes
into the one optimizer restriction which applies in concurrent environments,
which gets a bit more of an extended description because any optimization
dealing with stores needs to be aware of it.</p>
<p>From the optimizer's point of view, the rule is that if there
are not any instructions with atomic ordering involved, concurrency does
not matter, with one exception: if a variable might be visible to another
thread or signal handler, a store cannot be inserted along a path where it
might not execute otherwise. Take the following example:</p>
<pre>
/* C code, for readability; run through clang -O2 -S -emit-llvm to get
equivalent IR */
int x;
void f(int* a) {
for (int i = 0; i < 100; i++) {
if (a[i])
x += 1;
}
}
</pre>
<p>The following is equivalent in non-concurrent situations:</p>
<pre>
int x;
void f(int* a) {
int xtemp = x;
for (int i = 0; i < 100; i++) {
if (a[i])
xtemp += 1;
}
x = xtemp;
}
</pre>
<p>However, LLVM is not allowed to transform the former to the latter: it could
indirectly introduce undefined behavior if another thread can access x at
the same time. (This example is particularly of interest because before the
concurrency model was implemented, LLVM would perform this
transformation.)</p>
<p>Note that speculative loads are allowed; a load which
is part of a race returns <code>undef</code>, but does not have undefined
behavior.</p>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="atomicinst">Atomic instructions</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p>For cases where simple loads and stores are not sufficient, LLVM provides
various atomic instructions. The exact guarantees provided depend on the
ordering; see <a href="#ordering">Atomic orderings</a></p>
<p><code>load atomic</code> and <code>store atomic</code> provide the same
basic functionality as non-atomic loads and stores, but provide additional
guarantees in situations where threads and signals are involved.</p>
<p><code>cmpxchg</code> and <code>atomicrmw</code> are essentially like an
atomic load followed by an atomic store (where the store is conditional for
<code>cmpxchg</code>), but no other memory operation can happen on any thread
between the load and store. Note that LLVM's cmpxchg does not provide quite
as many options as the C++0x version.</p>
<p>A <code>fence</code> provides Acquire and/or Release ordering which is not
part of another operation; it is normally used along with Monotonic memory
operations. A Monotonic load followed by an Acquire fence is roughly
equivalent to an Acquire load.</p>
<p>Frontends generating atomic instructions generally need to be aware of the
target to some degree; atomic instructions are guaranteed to be lock-free,
and therefore an instruction which is wider than the target natively supports
can be impossible to generate.</p>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="ordering">Atomic orderings</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p>In order to achieve a balance between performance and necessary guarantees,
there are six levels of atomicity. They are listed in order of strength;
each level includes all the guarantees of the previous level except for
Acquire/Release. (See also <a href="LangRef.html#ordering">LangRef</a>.)</p>
<!-- ======================================================================= -->
<h3>
<a name="o_notatomic">NotAtomic</a>
</h3>
<div>
<p>NotAtomic is the obvious, a load or store which is not atomic. (This isn't
really a level of atomicity, but is listed here for comparison.) This is
essentially a regular load or store. If there is a race on a given memory
location, loads from that location return undef.</p>
<dl>
<dt>Relevant standard</dt>
<dd>This is intended to match shared variables in C/C++, and to be used
in any other context where memory access is necessary, and
a race is impossible. (The precise definition is in
<a href="LangRef.html#memmodel">LangRef</a>.)
<dt>Notes for frontends</dt>
<dd>The rule is essentially that all memory accessed with basic loads and
stores by multiple threads should be protected by a lock or other
synchronization; otherwise, you are likely to run into undefined
behavior. If your frontend is for a "safe" language like Java,
use Unordered to load and store any shared variable. Note that NotAtomic
volatile loads and stores are not properly atomic; do not try to use
them as a substitute. (Per the C/C++ standards, volatile does provide
some limited guarantees around asynchronous signals, but atomics are
generally a better solution.)
<dt>Notes for optimizers</dt>
<dd>Introducing loads to shared variables along a codepath where they would
not otherwise exist is allowed; introducing stores to shared variables
is not. See <a href="#outsideatomic">Optimization outside
atomic</a>.</dd>
<dt>Notes for code generation</dt>
<dd>The one interesting restriction here is that it is not allowed to write
to bytes outside of the bytes relevant to a store. This is mostly
relevant to unaligned stores: it is not allowed in general to convert
an unaligned store into two aligned stores of the same width as the
unaligned store. Backends are also expected to generate an i8 store
as an i8 store, and not an instruction which writes to surrounding
bytes. (If you are writing a backend for an architecture which cannot
satisfy these restrictions and cares about concurrency, please send an
email to llvmdev.)</dd>
</dl>
</div>
<!-- ======================================================================= -->
<h3>
<a name="o_unordered">Unordered</a>
</h3>
<div>
<p>Unordered is the lowest level of atomicity. It essentially guarantees that
races produce somewhat sane results instead of having undefined behavior.
It also guarantees the operation to be lock-free, so it do not depend on
the data being part of a special atomic structure or depend on a separate
per-process global lock. Note that code generation will fail for
unsupported atomic operations; if you need such an operation, use explicit
locking.</p>
<dl>
<dt>Relevant standard</dt>
<dd>This is intended to match the Java memory model for shared
variables.</dd>
<dt>Notes for frontends</dt>
<dd>This cannot be used for synchronization, but is useful for Java and
other "safe" languages which need to guarantee that the generated
code never exhibits undefined behavior. Note that this guarantee
is cheap on common platforms for loads of a native width, but can
be expensive or unavailable for wider loads, like a 64-bit store
on ARM. (A frontend for Java or other "safe" languages would normally
split a 64-bit store on ARM into two 32-bit unordered stores.)
<dt>Notes for optimizers</dt>
<dd>In terms of the optimizer, this prohibits any transformation that
transforms a single load into multiple loads, transforms a store
into multiple stores, narrows a store, or stores a value which
would not be stored otherwise. Some examples of unsafe optimizations
are narrowing an assignment into a bitfield, rematerializing
a load, and turning loads and stores into a memcpy call. Reordering
unordered operations is safe, though, and optimizers should take
advantage of that because unordered operations are common in
languages that need them.</dd>
<dt>Notes for code generation</dt>
<dd>These operations are required to be atomic in the sense that if you
use unordered loads and unordered stores, a load cannot see a value
which was never stored. A normal load or store instruction is usually
sufficient, but note that an unordered load or store cannot
be split into multiple instructions (or an instruction which
does multiple memory operations, like <code>LDRD</code> on ARM).</dd>
</dl>
</div>
<!-- ======================================================================= -->
<h3>
<a name="o_monotonic">Monotonic</a>
</h3>
<div>
<p>Monotonic is the weakest level of atomicity that can be used in
synchronization primitives, although it does not provide any general
synchronization. It essentially guarantees that if you take all the
operations affecting a specific address, a consistent ordering exists.
<dl>
<dt>Relevant standard</dt>
<dd>This corresponds to the C++0x/C1x <code>memory_order_relaxed</code>;
see those standards for the exact definition.
<dt>Notes for frontends</dt>
<dd>If you are writing a frontend which uses this directly, use with caution.
The guarantees in terms of synchronization are very weak, so make
sure these are only used in a pattern which you know is correct.
Generally, these would either be used for atomic operations which
do not protect other memory (like an atomic counter), or along with
a <code>fence</code>.</dd>
<dt>Notes for optimizers</dt>
<dd>In terms of the optimizer, this can be treated as a read+write on the
relevant memory location (and alias analysis will take advantage of
that). In addition, it is legal to reorder non-atomic and Unordered
loads around Monotonic loads. CSE/DSE and a few other optimizations
are allowed, but Monotonic operations are unlikely to be used in ways
which would make those optimizations useful.</dd>
<dt>Notes for code generation</dt>
<dd>Code generation is essentially the same as that for unordered for loads
and stores. No fences are required. <code>cmpxchg</code> and
<code>atomicrmw</code> are required to appear as a single operation.</dd>
</dl>
</div>
<!-- ======================================================================= -->
<h3>
<a name="o_acquire">Acquire</a>
</h3>
<div>
<p>Acquire provides a barrier of the sort necessary to acquire a lock to access
other memory with normal loads and stores.
<dl>
<dt>Relevant standard</dt>
<dd>This corresponds to the C++0x/C1x <code>memory_order_acquire</code>. It
should also be used for C++0x/C1x <code>memory_order_consume</code>.
<dt>Notes for frontends</dt>
<dd>If you are writing a frontend which uses this directly, use with caution.
Acquire only provides a semantic guarantee when paired with a Release
operation.</dd>
<dt>Notes for optimizers</dt>
<dd>Optimizers not aware of atomics can treat this like a nothrow call.
It is also possible to move stores from before an Acquire load
or read-modify-write operation to after it, and move non-Acquire
loads from before an Acquire operation to after it.</dd>
<dt>Notes for code generation</dt>
<dd>Architectures with weak memory ordering (essentially everything relevant
today except x86 and SPARC) require some sort of fence to maintain
the Acquire semantics. The precise fences required varies widely by
architecture, but for a simple implementation, most architectures provide
a barrier which is strong enough for everything (<code>dmb</code> on ARM,
<code>sync</code> on PowerPC, etc.). Putting such a fence after the
equivalent Monotonic operation is sufficient to maintain Acquire
semantics for a memory operation.</dd>
</dl>
</div>
<!-- ======================================================================= -->
<h3>
<a name="o_acquire">Release</a>
</h3>
<div>
<p>Release is similar to Acquire, but with a barrier of the sort necessary to
release a lock.
<dl>
<dt>Relevant standard</dt>
<dd>This corresponds to the C++0x/C1x <code>memory_order_release</code>.</dd>
<dt>Notes for frontends</dt>
<dd>If you are writing a frontend which uses this directly, use with caution.
Release only provides a semantic guarantee when paired with a Acquire
operation.</dd>
<dt>Notes for optimizers</dt>
<dd>Optimizers not aware of atomics can treat this like a nothrow call.
It is also possible to move loads from after a Release store
or read-modify-write operation to before it, and move non-Release
stores from after an Release operation to before it.</dd>
<dt>Notes for code generation</dt>
<dd>See the section on Acquire; a fence before the relevant operation is
usually sufficient for Release. Note that a store-store fence is not
sufficient to implement Release semantics; store-store fences are
generally not exposed to IR because they are extremely difficult to
use correctly.</dd>
</dl>
</div>
<!-- ======================================================================= -->
<h3>
<a name="o_acqrel">AcquireRelease</a>
</h3>
<div>
<p>AcquireRelease (<code>acq_rel</code> in IR) provides both an Acquire and a
Release barrier (for fences and operations which both read and write memory).
<dl>
<dt>Relevant standard</dt>
<dd>This corresponds to the C++0x/C1x <code>memory_order_acq_rel</code>.
<dt>Notes for frontends</dt>
<dd>If you are writing a frontend which uses this directly, use with caution.
Acquire only provides a semantic guarantee when paired with a Release
operation, and vice versa.</dd>
<dt>Notes for optimizers</dt>
<dd>In general, optimizers should treat this like a nothrow call; the
the possible optimizations are usually not interesting.</dd>
<dt>Notes for code generation</dt>
<dd>This operation has Acquire and Release semantics; see the sections on
Acquire and Release.</dd>
</dl>
</div>
<!-- ======================================================================= -->
<h3>
<a name="o_seqcst">SequentiallyConsistent</a>
</h3>
<div>
<p>SequentiallyConsistent (<code>seq_cst</code> in IR) provides
Acquire semantics for loads and Release semantics for
stores. Additionally, it guarantees that a total ordering exists
between all SequentiallyConsistent operations.
<dl>
<dt>Relevant standard</dt>
<dd>This corresponds to the C++0x/C1x <code>memory_order_seq_cst</code>,
Java volatile, and the gcc-compatible <code>__sync_*</code> builtins
which do not specify otherwise.
<dt>Notes for frontends</dt>
<dd>If a frontend is exposing atomic operations, these are much easier to
reason about for the programmer than other kinds of operations, and using
them is generally a practical performance tradeoff.</dd>
<dt>Notes for optimizers</dt>
<dd>Optimizers not aware of atomics can treat this like a nothrow call.
For SequentiallyConsistent loads and stores, the same reorderings are
allowed as for Acquire loads and Release stores, except that
SequentiallyConsistent operations may not be reordered.</dd>
<dt>Notes for code generation</dt>
<dd>SequentiallyConsistent loads minimally require the same barriers
as Acquire operations and SequeuentiallyConsistent stores require
Release barriers. Additionally, the code generator must enforce
ordering between SequeuentiallyConsistent stores followed by
SequeuentiallyConsistent loads. This is usually done by emitting
either a full fence before the loads or a full fence after the
stores; which is preferred varies by architecture.</dd>
</dl>
</div>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="iropt">Atomics and IR optimization</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p>Predicates for optimizer writers to query:
<ul>
<li>isSimple(): A load or store which is not volatile or atomic. This is
what, for example, memcpyopt would check for operations it might
transform.</li>
<li>isUnordered(): A load or store which is not volatile and at most
Unordered. This would be checked, for example, by LICM before hoisting
an operation.</li>
<li>mayReadFromMemory()/mayWriteToMemory(): Existing predicate, but note
that they return true for any operation which is volatile or at least
Monotonic.</li>
<li>Alias analysis: Note that AA will return ModRef for anything Acquire or
Release, and for the address accessed by any Monotonic operation.</li>
</ul>
<p>To support optimizing around atomic operations, make sure you are using
the right predicates; everything should work if that is done. If your
pass should optimize some atomic operations (Unordered operations in
particular), make sure it doesn't replace an atomic load or store with
a non-atomic operation.</p>
<p>Some examples of how optimizations interact with various kinds of atomic
operations:
<ul>
<li>memcpyopt: An atomic operation cannot be optimized into part of a
memcpy/memset, including unordered loads/stores. It can pull operations
across some atomic operations.
<li>LICM: Unordered loads/stores can be moved out of a loop. It just treats
monotonic operations like a read+write to a memory location, and anything
stricter than that like a nothrow call.
<li>DSE: Unordered stores can be DSE'ed like normal stores. Monotonic stores
can be DSE'ed in some cases, but it's tricky to reason about, and not
especially important.
<li>Folding a load: Any atomic load from a constant global can be
constant-folded, because it cannot be observed. Similar reasoning allows
scalarrepl with atomic loads and stores.
</ul>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="codegen">Atomics and Codegen</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p>Atomic operations are represented in the SelectionDAG with
<code>ATOMIC_*</code> opcodes. On architectures which use barrier
instructions for all atomic ordering (like ARM), appropriate fences are
split out as the DAG is built.</p>
<p>The MachineMemOperand for all atomic operations is currently marked as
volatile; this is not correct in the IR sense of volatile, but CodeGen
handles anything marked volatile very conservatively. This should get
fixed at some point.</p>
<p>Common architectures have some way of representing at least a pointer-sized
lock-free <code>cmpxchg</code>; such an operation can be used to implement
all the other atomic operations which can be represented in IR up to that
size. Backends are expected to implement all those operations, but not
operations which cannot be implemented in a lock-free manner. It is
expected that backends will give an error when given an operation which
cannot be implemented. (The LLVM code generator is not very helpful here
at the moment, but hopefully that will change.)</p>
<p>The implementation of atomics on LL/SC architectures (like ARM) is currently
a bit of a mess; there is a lot of copy-pasted code across targets, and
the representation is relatively unsuited to optimization (it would be nice
to be able to optimize loops involving cmpxchg etc.).</p>
<p>On x86, all atomic loads generate a <code>MOV</code>.
SequentiallyConsistent stores generate an <code>XCHG</code>, other stores
generate a <code>MOV</code>. SequentiallyConsistent fences generate an
<code>MFENCE</code>, other fences do not cause any code to be generated.
cmpxchg uses the <code>LOCK CMPXCHG</code> instruction.
<code>atomicrmw xchg</code> uses <code>XCHG</code>,
<code>atomicrmw add</code> and <code>atomicrmw sub</code> use
<code>XADD</code>, and all other <code>atomicrmw</code> operations generate
a loop with <code>LOCK CMPXCHG</code>. Depending on the users of the
result, some <code>atomicrmw</code> operations can be translated into
operations like <code>LOCK AND</code>, but that does not work in
general.</p>
<p>On ARM, MIPS, and many other RISC architectures, Acquire, Release, and
SequentiallyConsistent semantics require barrier instructions
for every such operation. Loads and stores generate normal instructions.
<code>cmpxchg</code> and <code>atomicrmw</code> can be represented using
a loop with LL/SC-style instructions which take some sort of exclusive
lock on a cache line (<code>LDREX</code> and <code>STREX</code> on
ARM, etc.). At the moment, the IR does not provide any way to represent a
weak <code>cmpxchg</code> which would not require a loop.</p>
</div>
<!-- *********************************************************************** -->
<hr>
<address>
<a href="http://jigsaw.w3.org/css-validator/check/referer"><img
src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
<a href="http://validator.w3.org/check/referer"><img
src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
<a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
Last modified: $Date: 2011-08-09 02:07:00 -0700 (Tue, 09 Aug 2011) $
</address>
</body>
</html>
|