aboutsummaryrefslogtreecommitdiffstats
path: root/lib/tools/doc/src/lcnt.xml
blob: 31e5c241e93a77cf541203195acb384f165f5849 (plain) (blame)
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
<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE erlref SYSTEM "erlref.dtd">

<erlref>
    <header>
	<copyright>
	    <year>2009</year>
	    <year>2017</year>
	    <holder>Ericsson AB, All Rights Reserved</holder>
	</copyright>
	<legalnotice>
      Licensed under the Apache License, Version 2.0 (the "License");
      you may not use this file except in compliance with the License.
      You may obtain a copy of the License at
 
          http://www.apache.org/licenses/LICENSE-2.0

      Unless required by applicable law or agreed to in writing, software
      distributed under the License is distributed on an "AS IS" BASIS,
      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
      See the License for the specific language governing permissions and
      limitations under the License.

      The Initial Developer of the Original Code is Ericsson AB.
	</legalnotice>

	<title>lcnt</title>
	<prepared>Björn-Egil Dahlberg</prepared>
	<responsible>nobody</responsible>
	<docno></docno>
	<approved>nobody</approved>
	<checked></checked>
	<date>2009-11-26</date>
	<rev>PA1</rev>
	<file>lcnt.xml</file>
    </header>
    <module>lcnt</module>
    <modulesummary>A runtime system Lock Profiling tool.</modulesummary>
    <description>
	<p>The <c>lcnt</c> module is used to profile the internal ethread locks in the
	    Erlang Runtime System. With <c>lcnt</c> enabled, internal counters in the
	    runtime system are updated each time a lock is taken. The counters stores
	    information about the number of acquisition tries and the number of collisions
	    that has occurred during the acquisition tries. The counters also record the
	    waiting time a lock has caused for a blocked thread when a collision has occurred.
	</p>
	<p>
	    The data produced by the lock counters will give an estimate on how well
	    the runtime system will behave from a parallelizable view point for the
	    scenarios tested. This tool was mainly developed to help Erlang runtime
	    developers iron out potential and generic bottlenecks.
	</p>
	<p>Locks in the emulator are named after what type of resource they protect and where
	    in the emulator they are initialized, those are lock 'classes'. Most of those
	    locks are also instantiated several times, and given unique identifiers, to increase
	    locking granularity. Typically an instantiated lock protects a disjunct set of
	    the resource, for example ets tables, processes or ports. In other cases it protects a
	    specific range of a resource, for example <c>pix_lock</c> which protects index to process
	    mappings, and is given a unique number within the class. A unique lock in <c>lcnt</c>
	    is referenced by a name (class) and an identifier: <c>{Name, Id}</c>.
	</p>
	<p>Some locks in the system are static and protects global resources, for example
	    <c>bif_timers</c> and the <c>run_queue</c> locks. Other locks are dynamic and not
	    necessarily long lived, for example process locks and ets-table locks. The
	    statistics data from short lived locks can be stored separately when the locks
	    are deleted. This behavior is by default turned off to save memory but can be
	    turned on via <c>lcnt:rt_opt({copy_save, true})</c>. The <c>lcnt:apply/1,2,3</c>
	    functions enables this behavior during profiling.
	</p>
    </description>
    <funcs>

	<func>
	    <name>start() -> {ok, Pid} | {error, {already_started, Pid}} </name>
	    <fsummary>Starts the lock profiler server.</fsummary>
	    <type>
		<v>Pid = pid()</v>
	    </type>
	    <desc>
		<p>Starts the lock profiler server. The server only act as a medium for the
		    user and performs filtering and printing of data collected by <c>lcnt:collect/1</c>.
		</p>
	    </desc>
	</func>

	<func>
	    <name>stop() -> ok</name>
	    <fsummary>Stops the lock profiler server.</fsummary>
	    <desc>
		<p>Stops the lock profiler server.</p>
	    </desc>
	</func>

	<func>
	    <name>collect() -> ok</name>
	    <fsummary>Same as <c>collect(node())</c>.</fsummary>
	    <desc><p>Same as <c>collect(node())</c>.</p></desc>
	</func>

	<func>
	    <name>collect(Node) -> ok</name>
	    <fsummary>Collects lock statistics from the runtime system.</fsummary>
	    <type>
		<v>Node = node()</v>
	    </type>
	    <desc>
		<p>Collects lock statistics from the runtime system. The function starts a
		    server if it is not already started. It then populates the server with lock
		    statistics. If the server held any lock statistics data before the collect then
		    that data is lost.
		</p>
		<note>
		    <p>
			When collection occurs the runtime system transitions to a single thread,
			blocking all other threads. No other tasks will be scheduled during this
			operation. Depending on the size of the data this might take a long time
			(several seconds) and cause timeouts in the system.
		    </p>
		</note>
	    </desc>
	</func>

	<func>
	    <name>clear() -> ok</name>
	    <fsummary>Same as <c>clear(node())</c>.</fsummary>
	    <desc><p>Same as <c>clear(node())</c>.</p></desc>
	</func>

	<func>
	    <name>clear(Node) -> ok</name>
	    <fsummary>Clears the internal lock statistics from runtime system.</fsummary>
	    <type>
		<v>Node = node()</v>
	    </type>
	    <desc>
		<p>Clears the internal lock statistics from the runtime system. This does not clear the
		    data on the server only on runtime system. All counters for static locks are zeroed,
		    all dynamic locks currently alive are zeroed and all saved locks now destroyed are removed.
		    It also resets the duration timer.
		</p>
	    </desc>
	</func>
	<func>
	    <name>conflicts() -> ok</name>
	    <fsummary>Same as <c>conflicts([])</c>.</fsummary>
	    <desc><p>Same as <c>conflicts([])</c>.</p></desc>
	</func>
	<func>
	    <name>conflicts([Option]) -> ok</name>
	    <fsummary>Prints a list of internal lock counters.</fsummary>
	    <type>
		<v>Option     = {sort, Sort} | {reverse, bool()} | {thresholds, [Thresholds]} | {print, [Print | {Print, integer()}]} | {max_locks, MaxLocks} | {combine, bool()}</v>
		<v>Sort       = name | id | type | tries | colls | ratio | time | entry</v>
		<v>Thresholds = {tries, integer()} | {colls, integer()} | {time, integer()}</v>
		<v>Print      = name | id | type | entry | tries | colls | ratio | time | duration</v>
		<v>MaxLocks   = integer() | none</v>
	    </type>
	    <desc>
		<p>Prints a list of internal locks and its statistics.</p>
		<p>For option description, see <seealso marker="#inspect/2">lcnt:inspect/2</seealso>.</p>
	    </desc>
	</func>

	<func>
	    <name>locations() -> ok</name>
	    <fsummary>Same as <c>locations([])</c>.</fsummary>
	    <desc>
		<p>Same as <c>locations([])</c>.</p>
	    </desc>
	</func>
	<func>
	    <name>locations([Option]) -> ok</name>
	    <fsummary>Prints a list of internal lock counters by source code locations.</fsummary>
	    <type>
		<v>Option     = {sort, Sort} | {thresholds, [Thresholds]} | {print, [Print | {Print, integer()}]} | {max_locks, MaxLocks} | {combine, bool()}</v>
		<v>Sort       = name | id | type | tries | colls | ratio | time | entry</v>
		<v>Thresholds = {tries, integer()} | {colls, integer()} | {time, integer()}</v>
		<v>Print      = name | id | type | entry | tries | colls | ratio | time | duration</v>
		<v>MaxLocks   = integer() | none</v>
	    </type>
	    <desc>
		<p>Prints a list of internal lock counters by source code locations.</p>
		<p>For option description, see <seealso marker="#inspect/2">lcnt:inspect/2</seealso>.</p>
	    </desc>
	</func>

	<func>
	    <name>inspect(Lock) -> ok</name>
	    <fsummary>Same as <c>inspect(Lock, [])</c>.</fsummary>
	    <desc><p>Same as <c>inspect(Lock, [])</c>.</p></desc>
	</func>
	<func>
	    <name>inspect(Lock, [Option]) -> ok</name>
	    <fsummary>Prints a list of internal lock counters for a specific lock.</fsummary>
	    <type>
		<v>Lock       = Name | {Name, Id | [Id]}</v>
		<v>Name       = atom() | pid() | port()</v>
		<v>Id         = atom() | integer() | pid() | port()</v>
		<v>Option     = {sort, Sort} | {thresholds, [Thresholds]} | {print, [Print | {Print, integer()}]} | {max_locks, MaxLocks} | {combine, bool()} | {locations, bool()}</v>
		<v>Sort       = name | id | type | tries | colls | ratio | time</v>
		<v>Thresholds = {tries, integer()} | {colls, integer()} | {time, integer()}</v>
		<v>Print      = name | id | type | entry | tries | colls | ratio | time | duration</v>
		<v>MaxLocks   = integer() | none</v>
	    </type>
	    <desc>
		<p>Prints a list of internal lock counters for a specific lock.</p>
		<p>Lock <c>Name</c> and <c>Id</c> for ports and processes are interchangeable with the use of <c>lcnt:swap_pid_keys/0</c> and is the reason why <c>pid()</c> and <c>port()</c> options can be used in both <c>Name</c> and <c>Id</c> space. Both pids and ports are special identifiers with stripped creation and can be recreated with <seealso marker="#pid/3">lcnt:pid/2,3</seealso> and <seealso marker="#port/2">lcnt:port/1,2</seealso>.  </p>
		<p>Option description:</p>
		<taglist>
		    <tag><c>{combine, bool()}</c></tag>
		    <item>Combine the statistics from different instances of a lock class.
			<br/>Default: <c>true</c>
		    </item>

		    <tag><c>{locations, bool()}</c></tag>
		    <item>Print the statistics by source file and line numbers.
			<br/>Default: <c>false</c>
		    </item>

		    <tag><c>{max_locks, MaxLocks}</c></tag>
		    <item>Maximum number of locks printed or no limit with <c>none</c>.
			<br/>Default: <c>20</c>
		    </item>

		    <tag><c>{print, PrintOptions}</c></tag>
		    <item>Printing options:
			<taglist>
			    <tag><c>name</c></tag>
			    <item>Named lock or named set of locks (classes). The same name used for initializing the lock in the VM.</item>

			    <tag><c>id</c></tag>
			    <item>Internal id for set of locks, not always unique. This could be table name for ets tables (db_tab), port id for ports, integer identifiers for allocators, etc.</item>

			    <tag><c>type</c></tag>
			    <item>Type of lock: <c>rw_mutex</c>, <c>mutex</c>, <c>spinlock</c>, <c>rw_spinlock</c> or <c>proclock</c>.</item>

			    <tag><c>entry</c></tag>
			    <item>In combination with <c>{locations, true}</c> this option prints the lock operations source file and line number entry-points along with statistics for each entry. </item>

			    <tag><c>tries</c></tag>
			    <item>Number of acquisitions of this lock.</item>

			    <tag><c>colls</c></tag>
			    <item>Number of collisions when a thread tried to acquire this lock. This is when a trylock is EBUSY, a write try on read held rw_lock, a try read on write held rw_lock, a thread tries to lock an already locked lock. (Internal states supervises this).</item>

			    <tag><c>ratio</c></tag>
			    <item>The ratio between the number of collisions and the number of tries (acquisitions) in percentage.</item>

			    <tag><c>time</c></tag>
			    <item>Accumulated waiting time for this lock. This could be greater than actual wall clock time, it is accumulated for all threads. Trylock conflicts does not accumulate time.</item>

			    <tag><c>duration</c></tag>
			    <item>Percentage of accumulated waiting time of wall clock time. This percentage can be higher than 100% since accumulated time is from all threads.</item>
			</taglist>
			<br/>Default: <c>[name,id,tries,colls,ratio,time,duration]</c>
		    </item>

		    <tag><c>{reverse, bool()}</c></tag>
		    <item>Reverses the order of sorting.
			<br/>Default: <c>false</c>
		    </item>

		    <tag><c>{sort, Sort}</c></tag>
		    <item>Column sorting orders.
			<br/>Default: <c>time</c>
		    </item>

		    <tag><c>{thresholds, Thresholds}</c></tag>
		    <item>Filtering thresholds. Anything values above the threshold value are passed through.
			<br/>Default: <c>[{tries, 0}, {colls, 0}, {time, 0}]</c>
		    </item>

		</taglist>

	    </desc>
	</func>

	<func>
	    <name>information() -> ok</name>
	    <fsummary>Prints lcnt server state and generic information about collected lock statistics.</fsummary>
	    <desc>
		<p>Prints lcnt server state and generic information about collected lock statistics.</p>
	    </desc>
	</func>

	<func>
	    <name>swap_pid_keys() -> ok</name>
	    <fsummary>Swaps places on <c>Name</c> and <c>Id</c> space for ports and processes.</fsummary>
	    <desc>
		<p>Swaps places on <c>Name</c> and <c>Id</c> space for ports and processes.</p>
	    </desc>
	</func>

	<func>
	    <name>load(Filename) -> ok</name>
	    <fsummary>Restores previously saved data to the server.</fsummary>
	    <type>
		<v>Filename = filename()</v>
	    </type>
	    <desc>
		<p>Restores previously saved data to the server.</p>
	    </desc>
	</func>

	<func>
	    <name>save(Filename) -> ok</name>
	    <fsummary>Saves the collected data to file.</fsummary>
	    <type>
		<v>Filename = filename()</v>
	    </type>
	    <desc>
		<p>Saves the collected data to file.</p>
	    </desc>
	</func>
    </funcs>

    <section>
	<title>Convenience functions</title>
	<p>The following functions are used for convenience.</p>
    </section>
    <funcs>
	<func>
	    <name>apply(Fun) -> term()</name>
	    <fsummary>Same as <c>apply(Fun, [])</c>.</fsummary>
	    <desc>
		<p>Same as <c>apply(Fun, [])</c>.</p>
	    </desc>
	</func>
	<func>
	    <name>apply(Fun, Args) -> term()</name>
	    <fsummary>Clears counters, applies function and collects the profiling results.</fsummary>
	    <type>
		<v>Fun = fun()</v>
		<v>Args = [term()]</v>
	    </type>
	    <desc>
		<p> Clears the lock counters and then setups the instrumentation to save all destroyed locks.
		    After setup the fun is called, passing the elements in <c>Args</c> as arguments.
		    When the fun returns the statistics are immediately collected to the server. After the
		    collection the instrumentation is returned to its previous behavior.
		    The result of the applied fun is returned.
		</p>
	    </desc>
	</func>
	<func>
	    <name>apply(Module, Function, Args) -> term()</name>
	    <fsummary>Clears counters, applies function and collects the profiling results.</fsummary>
	    <type>
		<v>Module = atom()</v>
		<v>Function = atom()</v>
		<v>Args = [term()]</v>
	    </type>
	    <desc>
		<p> Clears the lock counters and then setups the instrumentation to save all destroyed locks.
		    After setup the function is called, passing the elements in <c>Args</c> as arguments.
		    When the function returns the statistics are immediately collected to the server. After the
		    collection the instrumentation is returned to its previous behavior.
		    The result of the applied function is returned.
		</p>
	    </desc>
	</func>

	<func>
	    <name>pid(Id, Serial) -> pid()</name>
	    <fsummary>Same as <c>pid(node(), Id, Serial)</c>.</fsummary>
	    <desc><p>Same as <c>pid(node(), Id, Serial)</c>.</p></desc>
	</func>
	<func>
	    <name>pid(Node, Id, Serial) -> pid()</name>
	    <fsummary>Creates a process id with creation 0.</fsummary>
	    <type>
		<v>Node = node()</v>
		<v>Id = integer()</v>
		<v>Serial = integer()</v>
	    </type>
	    <desc>
		<p>Creates a process id with creation 0. Example:</p>
	    </desc>
	</func>

	<func>
	    <name>port(Id) -> port()</name>
	    <fsummary>Same as <c>port(node(), Id)</c>.</fsummary>
	    <desc><p>Same as <c>port(node(), Id)</c>.</p></desc>
	</func>
	<func>
	    <name>port(Node, Id) -> port()</name>
	    <fsummary>Creates a port id with creation 0.</fsummary>
	    <type>
		<v>Node = node()</v>
		<v>Id = integer()</v>
	    </type>
	    <desc><p>Creates a port id with creation 0.</p></desc>
	</func>

    </funcs>

    <section>
	<title>Internal runtime lock counter controllers</title>
	<p> The following functions control the behavior of the internal counters.  </p>
    </section>

    <funcs>
	<func>
	    <name>rt_collect() -> [lock_counter_data()]</name>
	    <fsummary>Same as <c>rt_collect(node())</c>.</fsummary>
	    <desc> <p>Same as <c>rt_collect(node())</c>.</p> </desc>
	</func>
	<func>
	    <name>rt_collect(Node) -> [lock_counter_data()]</name>
	    <fsummary>Returns a list of raw lock counter data.</fsummary>
	    <type>
		<v>Node = node()</v>
	    </type>
	    <desc> <p>Returns a list of raw lock counter data.</p> </desc>
	</func>

	<func>
	    <name>rt_clear() -> ok</name>
	    <fsummary>Same as <c>rt_clear(node())</c>.</fsummary>
	    <desc> <p>Same as <c>rt_clear(node())</c>.</p> </desc>
	</func>
	<func>
	    <name>rt_clear(Node) -> ok</name>
	    <fsummary>Clears the internal counters.</fsummary>
	    <type>
		<v>Node = node()</v>
	    </type>
	    <desc> <p>Clear the internal counters. Same as <c>lcnt:clear(Node)</c>.</p></desc>
	</func>

	<func>
	    <name>rt_opt({Type, bool()}) -> bool()</name>
	    <fsummary>Same as <c>rt_opt(node(), {Type, Opt})</c>.</fsummary>
	    <desc> <p>Same as <c>rt_opt(node(), {Type, Opt})</c>.</p> </desc>
	</func>
	<func>
	    <name>rt_opt(Node, {Type, bool()}) -> bool()</name>
	    <fsummary>Changes the lock counter behavior and returns the previous behaviour.</fsummary>
	    <type>
		<v>Node = node()</v>
		<v>Type = copy_save | process_locks</v>
	    </type>
	    <desc>
		<p>Changes the lock counter behavior and returns the previous behaviour.</p>
		<p>Option description:</p>
		<taglist>
		    <tag><c>{copy_save, bool()}</c></tag>
		    <item>Enable statistics saving from destroyed locks by copying. This might consume a lot of memory.
			<br/>Default: <c>false</c>
		    </item>

		    <tag><c>{process_locks, bool()}</c></tag>
		    <item>Profile process locks.
			<br/>Default: <c>true</c>
		    </item>
		</taglist>
	    </desc>
	</func>
    </funcs>

    <section>
	<title>See Also</title>
	<p> <seealso marker="lcnt_chapter">LCNT User's Guide</seealso></p>
    </section>
</erlref>