Document new GC options -q1 and -qg<n>
[ghc.git] / docs / users_guide / runtime_control.xml
1 <?xml version="1.0" encoding="iso-8859-1"?>
2 <sect1 id="runtime-control">
3 <title>Running a compiled program</title>
4
5 <indexterm><primary>runtime control of Haskell programs</primary></indexterm>
6 <indexterm><primary>running, compiled program</primary></indexterm>
7 <indexterm><primary>RTS options</primary></indexterm>
8
9 <para>To make an executable program, the GHC system compiles your
10 code and then links it with a non-trivial runtime system (RTS),
11 which handles storage management, profiling, etc.</para>
12
13 <para>You have some control over the behaviour of the RTS, by giving
14 special command-line arguments to your program.</para>
15
16 <para>When your Haskell program starts up, its RTS extracts
17 command-line arguments bracketed between
18 <option>+RTS</option><indexterm><primary><option>+RTS</option></primary></indexterm>
19 and
20 <option>-RTS</option><indexterm><primary><option>-RTS</option></primary></indexterm>
21 as its own. For example:</para>
22
23 <screen>
24 % ./a.out -f +RTS -p -S -RTS -h foo bar
25 </screen>
26
27 <para>The RTS will snaffle <option>-p</option> <option>-S</option>
28 for itself, and the remaining arguments <literal>-f -h foo bar</literal>
29 will be handed to your program if/when it calls
30 <function>System.getArgs</function>.</para>
31
32 <para>No <option>-RTS</option> option is required if the
33 runtime-system options extend to the end of the command line, as in
34 this example:</para>
35
36 <screen>
37 % hls -ltr /usr/etc +RTS -A5m
38 </screen>
39
40 <para>If you absolutely positively want all the rest of the options
41 in a command line to go to the program (and not the RTS), use a
42 <option>&ndash;&ndash;RTS</option><indexterm><primary><option>--RTS</option></primary></indexterm>.</para>
43
44 <para>As always, for RTS options that take
45 <replaceable>size</replaceable>s: If the last character of
46 <replaceable>size</replaceable> is a K or k, multiply by 1000; if an
47 M or m, by 1,000,000; if a G or G, by 1,000,000,000. (And any
48 wraparound in the counters is <emphasis>your</emphasis>
49 fault!)</para>
50
51 <para>Giving a <literal>+RTS -f</literal>
52 <indexterm><primary><option>-f</option></primary><secondary>RTS option</secondary></indexterm> option
53 will print out the RTS options actually available in your program
54 (which vary, depending on how you compiled).</para>
55
56 <para>NOTE: since GHC is itself compiled by GHC, you can change RTS
57 options in the compiler using the normal
58 <literal>+RTS ... -RTS</literal>
59 combination. eg. to increase the maximum heap
60 size for a compilation to 128M, you would add
61 <literal>+RTS -M128m -RTS</literal>
62 to the command line.</para>
63
64 <sect2 id="rts-optinos-environment">
65 <title>Setting global RTS options</title>
66
67 <indexterm><primary>RTS options</primary><secondary>from the environment</secondary></indexterm>
68 <indexterm><primary>environment variable</primary><secondary>for
69 setting RTS options</secondary></indexterm>
70
71 <para>RTS options are also taken from the environment variable
72 <envar>GHCRTS</envar><indexterm><primary><envar>GHCRTS</envar></primary>
73 </indexterm>. For example, to set the maximum heap size
74 to 128M for all GHC-compiled programs (using an
75 <literal>sh</literal>-like shell):</para>
76
77 <screen>
78 GHCRTS='-M128m'
79 export GHCRTS
80 </screen>
81
82 <para>RTS options taken from the <envar>GHCRTS</envar> environment
83 variable can be overridden by options given on the command
84 line.</para>
85
86 </sect2>
87
88 <sect2 id="rts-options-misc">
89 <title>Miscellaneous RTS options</title>
90
91 <variablelist>
92 <varlistentry>
93 <term><option>-V<replaceable>secs</replaceable></option>
94 <indexterm><primary><option>-V</option></primary><secondary>RTS
95 option</secondary></indexterm></term>
96 <listitem>
97 <para>Sets the interval that the RTS clock ticks at. The
98 runtime uses a single timer signal to count ticks; this timer
99 signal is used to control the context switch timer (<xref
100 linkend="using-concurrent" />) and the heap profiling
101 timer <xref linkend="rts-options-heap-prof" />. Also, the
102 time profiler uses the RTS timer signal directly to record
103 time profiling samples.</para>
104
105 <para>Normally, setting the <option>-V</option> option
106 directly is not necessary: the resolution of the RTS timer is
107 adjusted automatically if a short interval is requested with
108 the <option>-C</option> or <option>-i</option> options.
109 However, setting <option>-V</option> is required in order to
110 increase the resolution of the time profiler.</para>
111
112 <para>Using a value of zero disables the RTS clock
113 completely, and has the effect of disabling timers that
114 depend on it: the context switch timer and the heap profiling
115 timer. Context switches will still happen, but
116 deterministically and at a rate much faster than normal.
117 Disabling the interval timer is useful for debugging, because
118 it eliminates a source of non-determinism at runtime.</para>
119 </listitem>
120 </varlistentry>
121
122 <varlistentry>
123 <term><option>--install-signal-handlers=<replaceable>yes|no</replaceable></option>
124 <indexterm><primary><option>--install-signal-handlers</option></primary><secondary>RTS
125 option</secondary></indexterm></term>
126 <listitem>
127 <para>If yes (the default), the RTS installs signal handlers to catch
128 things like ctrl-C. This option is primarily useful for when
129 you are using the Haskell code as a DLL, and want to set your
130 own signal handlers.</para>
131 </listitem>
132 </varlistentry>
133
134 <varlistentry>
135 <term><option>-xm<replaceable>address</replaceable></option>
136 <indexterm><primary><option>-xm</option></primary><secondary>RTS
137 option</secondary></indexterm></term>
138 <listitem>
139 <para>
140 WARNING: this option is for working around memory
141 allocation problems only. Do not use unless GHCi fails
142 with a message like &ldquo;<literal>failed to mmap() memory below 2Gb</literal>&rdquo;. If you need to use this option to get GHCi working
143 on your machine, please file a bug.
144 </para>
145
146 <para>
147 On 64-bit machines, the RTS needs to allocate memory in the
148 low 2Gb of the address space. Support for this across
149 different operating systems is patchy, and sometimes fails.
150 This option is there to give the RTS a hint about where it
151 should be able to allocate memory in the low 2Gb of the
152 address space. For example, <literal>+RTS -xm20000000
153 -RTS</literal> would hint that the RTS should allocate
154 starting at the 0.5Gb mark. The default is to use the OS's
155 built-in support for allocating memory in the low 2Gb if
156 available (e.g. <literal>mmap</literal>
157 with <literal>MAP_32BIT</literal> on Linux), or
158 otherwise <literal>-xm40000000</literal>.
159 </para>
160 </listitem>
161 </varlistentry>
162 </variablelist>
163 </sect2>
164
165 <sect2 id="rts-options-gc">
166 <title>RTS options to control the garbage collector</title>
167
168 <indexterm><primary>garbage collector</primary><secondary>options</secondary></indexterm>
169 <indexterm><primary>RTS options</primary><secondary>garbage collection</secondary></indexterm>
170
171 <para>There are several options to give you precise control over
172 garbage collection. Hopefully, you won't need any of these in
173 normal operation, but there are several things that can be tweaked
174 for maximum performance.</para>
175
176 <variablelist>
177
178 <varlistentry>
179 <term>
180 <option>-A</option><replaceable>size</replaceable>
181 <indexterm><primary><option>-A</option></primary><secondary>RTS option</secondary></indexterm>
182 <indexterm><primary>allocation area, size</primary></indexterm>
183 </term>
184 <listitem>
185 <para>&lsqb;Default: 256k&rsqb; Set the allocation area size
186 used by the garbage collector. The allocation area
187 (actually generation 0 step 0) is fixed and is never resized
188 (unless you use <option>-H</option>, below).</para>
189
190 <para>Increasing the allocation area size may or may not
191 give better performance (a bigger allocation area means
192 worse cache behaviour but fewer garbage collections and less
193 promotion).</para>
194
195 <para>With only 1 generation (<option>-G1</option>) the
196 <option>-A</option> option specifies the minimum allocation
197 area, since the actual size of the allocation area will be
198 resized according to the amount of data in the heap (see
199 <option>-F</option>, below).</para>
200 </listitem>
201 </varlistentry>
202
203 <varlistentry>
204 <term>
205 <option>-c</option>
206 <indexterm><primary><option>-c</option></primary><secondary>RTS option</secondary></indexterm>
207 <indexterm><primary>garbage collection</primary><secondary>compacting</secondary></indexterm>
208 <indexterm><primary>compacting garbage collection</primary></indexterm>
209 </term>
210 <listitem>
211 <para>Use a compacting algorithm for collecting the oldest
212 generation. By default, the oldest generation is collected
213 using a copying algorithm; this option causes it to be
214 compacted in-place instead. The compaction algorithm is
215 slower than the copying algorithm, but the savings in memory
216 use can be considerable.</para>
217
218 <para>For a given heap size (using the <option>-H</option>
219 option), compaction can in fact reduce the GC cost by
220 allowing fewer GCs to be performed. This is more likely
221 when the ratio of live data to heap size is high, say
222 &gt;30&percnt;.</para>
223
224 <para>NOTE: compaction doesn't currently work when a single
225 generation is requested using the <option>-G1</option>
226 option.</para>
227 </listitem>
228 </varlistentry>
229
230 <varlistentry>
231 <term><option>-c</option><replaceable>n</replaceable></term>
232
233 <listitem>
234 <para>&lsqb;Default: 30&rsqb; Automatically enable
235 compacting collection when the live data exceeds
236 <replaceable>n</replaceable>&percnt; of the maximum heap size
237 (see the <option>-M</option> option). Note that the maximum
238 heap size is unlimited by default, so this option has no
239 effect unless the maximum heap size is set with
240 <option>-M</option><replaceable>size</replaceable>. </para>
241 </listitem>
242 </varlistentry>
243
244 <varlistentry>
245 <term>
246 <option>-F</option><replaceable>factor</replaceable>
247 <indexterm><primary><option>-F</option></primary><secondary>RTS option</secondary></indexterm>
248 <indexterm><primary>heap size, factor</primary></indexterm>
249 </term>
250 <listitem>
251
252 <para>&lsqb;Default: 2&rsqb; This option controls the amount
253 of memory reserved for the older generations (and in the
254 case of a two space collector the size of the allocation
255 area) as a factor of the amount of live data. For example,
256 if there was 2M of live data in the oldest generation when
257 we last collected it, then by default we'll wait until it
258 grows to 4M before collecting it again.</para>
259
260 <para>The default seems to work well here. If you have
261 plenty of memory, it is usually better to use
262 <option>-H</option><replaceable>size</replaceable> than to
263 increase
264 <option>-F</option><replaceable>factor</replaceable>.</para>
265
266 <para>The <option>-F</option> setting will be automatically
267 reduced by the garbage collector when the maximum heap size
268 (the <option>-M</option><replaceable>size</replaceable>
269 setting) is approaching.</para>
270 </listitem>
271 </varlistentry>
272
273 <varlistentry>
274 <term>
275 <option>-G</option><replaceable>generations</replaceable>
276 <indexterm><primary><option>-G</option></primary><secondary>RTS option</secondary></indexterm>
277 <indexterm><primary>generations, number of</primary></indexterm>
278 </term>
279 <listitem>
280 <para>&lsqb;Default: 2&rsqb; Set the number of generations
281 used by the garbage collector. The default of 2 seems to be
282 good, but the garbage collector can support any number of
283 generations. Anything larger than about 4 is probably not a
284 good idea unless your program runs for a
285 <emphasis>long</emphasis> time, because the oldest
286 generation will hardly ever get collected.</para>
287
288 <para>Specifying 1 generation with <option>+RTS -G1</option>
289 gives you a simple 2-space collector, as you would expect.
290 In a 2-space collector, the <option>-A</option> option (see
291 above) specifies the <emphasis>minimum</emphasis> allocation
292 area size, since the allocation area will grow with the
293 amount of live data in the heap. In a multi-generational
294 collector the allocation area is a fixed size (unless you
295 use the <option>-H</option> option, see below).</para>
296 </listitem>
297 </varlistentry>
298
299 <varlistentry>
300 <term>
301 <option>-q1</option>
302 <indexterm><primary><option>-q1</option><secondary>RTS
303 option</secondary></primary></indexterm>
304 </term>
305 <listitem>
306 <para>&lsqb;New in GHC 6.12.1&rsqb; Disable the parallel GC.
307 The parallel GC is turned on automatically when parallel
308 execution is enabled with the <option>-N</option> option;
309 this option is available to turn it off if
310 necessary.</para>
311
312 <para>Experiments have shown that parallel GC usually
313 results in a performance improvement given 3 cores or
314 more; with 2 cores it may or may not be beneficial,
315 depending on the workload. Bigger heaps work better with
316 parallel GC, so set your <option>-H</option> value high (3
317 or more times the maximum residency). Look at the timing
318 stats with <option>+RTS -s</option> to see whether you're
319 getting any benefit from parallel GC or not. If you find
320 parallel GC is significantly <emphasis>slower</emphasis>
321 (in elapsed time) than sequential GC, please report it as
322 a bug.</para>
323
324 <para>In GHC 6.10.1 it was possible to use a different
325 number of threads for GC than for execution, because the GC
326 used its own pool of threads. Now, the GC uses the same
327 threads as the mutator (for executing the program).</para>
328 </listitem>
329 </varlistentry>
330
331 <varlistentry>
332 <term>
333 <option>-qg<replaceable>n</replaceable></option>
334 <indexterm><primary><option>-qg</option><secondary>RTS
335 option</secondary></primary></indexterm>
336 </term>
337 <listitem>
338 <para>
339 &lsqb;Default: 1&rsqb; &lsqb;New in GHC 6.12.1&rsqb;
340 Enable the parallel GC only in
341 generation <replaceable>n</replaceable> and greater.
342 Parallel GC is often not worthwhile for collections in
343 generation 0 (the young generation), so it is enabled by
344 default only for collections in generation 1 (and higher,
345 if applicable).
346 </para>
347 </listitem>
348 </varlistentry>
349
350 <varlistentry>
351 <term>
352 <option>-H</option><replaceable>size</replaceable>
353 <indexterm><primary><option>-H</option></primary><secondary>RTS option</secondary></indexterm>
354 <indexterm><primary>heap size, suggested</primary></indexterm>
355 </term>
356 <listitem>
357 <para>&lsqb;Default: 0&rsqb; This option provides a
358 &ldquo;suggested heap size&rdquo; for the garbage collector. The
359 garbage collector will use about this much memory until the
360 program residency grows and the heap size needs to be
361 expanded to retain reasonable performance.</para>
362
363 <para>By default, the heap will start small, and grow and
364 shrink as necessary. This can be bad for performance, so if
365 you have plenty of memory it's worthwhile supplying a big
366 <option>-H</option><replaceable>size</replaceable>. For
367 improving GC performance, using
368 <option>-H</option><replaceable>size</replaceable> is
369 usually a better bet than
370 <option>-A</option><replaceable>size</replaceable>.</para>
371 </listitem>
372 </varlistentry>
373
374 <varlistentry>
375 <term>
376 <option>-I</option><replaceable>seconds</replaceable>
377 <indexterm><primary><option>-I</option></primary>
378 <secondary>RTS option</secondary>
379 </indexterm>
380 <indexterm><primary>idle GC</primary>
381 </indexterm>
382 </term>
383 <listitem>
384 <para>(default: 0.3) In the threaded and SMP versions of the RTS (see
385 <option>-threaded</option>, <xref linkend="options-linker" />), a
386 major GC is automatically performed if the runtime has been idle
387 (no Haskell computation has been running) for a period of time.
388 The amount of idle time which must pass before a GC is performed is
389 set by the <option>-I</option><replaceable>seconds</replaceable>
390 option. Specifying <option>-I0</option> disables the idle GC.</para>
391
392 <para>For an interactive application, it is probably a good idea to
393 use the idle GC, because this will allow finalizers to run and
394 deadlocked threads to be detected in the idle time when no Haskell
395 computation is happening. Also, it will mean that a GC is less
396 likely to happen when the application is busy, and so
397 responsiveness may be improved. However, if the amount of live data in
398 the heap is particularly large, then the idle GC can cause a
399 significant delay, and too small an interval could adversely affect
400 interactive responsiveness.</para>
401
402 <para>This is an experimental feature, please let us know if it
403 causes problems and/or could benefit from further tuning.</para>
404 </listitem>
405 </varlistentry>
406
407 <varlistentry>
408 <term>
409 <option>-k</option><replaceable>size</replaceable>
410 <indexterm><primary><option>-k</option></primary><secondary>RTS option</secondary></indexterm>
411 <indexterm><primary>stack, minimum size</primary></indexterm>
412 </term>
413 <listitem>
414 <para>&lsqb;Default: 1k&rsqb; Set the initial stack size for
415 new threads. Thread stacks (including the main thread's
416 stack) live on the heap, and grow as required. The default
417 value is good for concurrent applications with lots of small
418 threads; if your program doesn't fit this model then
419 increasing this option may help performance.</para>
420
421 <para>The main thread is normally started with a slightly
422 larger heap to cut down on unnecessary stack growth while
423 the program is starting up.</para>
424 </listitem>
425 </varlistentry>
426
427 <varlistentry>
428 <term>
429 <option>-K</option><replaceable>size</replaceable>
430 <indexterm><primary><option>-K</option></primary><secondary>RTS option</secondary></indexterm>
431 <indexterm><primary>stack, maximum size</primary></indexterm>
432 </term>
433 <listitem>
434 <para>&lsqb;Default: 8M&rsqb; Set the maximum stack size for
435 an individual thread to <replaceable>size</replaceable>
436 bytes. This option is there purely to stop the program
437 eating up all the available memory in the machine if it gets
438 into an infinite loop.</para>
439 </listitem>
440 </varlistentry>
441
442 <varlistentry>
443 <term>
444 <option>-m</option><replaceable>n</replaceable>
445 <indexterm><primary><option>-m</option></primary><secondary>RTS option</secondary></indexterm>
446 <indexterm><primary>heap, minimum free</primary></indexterm>
447 </term>
448 <listitem>
449 <para>Minimum &percnt; <replaceable>n</replaceable> of heap
450 which must be available for allocation. The default is
451 3&percnt;.</para>
452 </listitem>
453 </varlistentry>
454
455 <varlistentry>
456 <term>
457 <option>-M</option><replaceable>size</replaceable>
458 <indexterm><primary><option>-M</option></primary><secondary>RTS option</secondary></indexterm>
459 <indexterm><primary>heap size, maximum</primary></indexterm>
460 </term>
461 <listitem>
462 <para>&lsqb;Default: unlimited&rsqb; Set the maximum heap size to
463 <replaceable>size</replaceable> bytes. The heap normally
464 grows and shrinks according to the memory requirements of
465 the program. The only reason for having this option is to
466 stop the heap growing without bound and filling up all the
467 available swap space, which at the least will result in the
468 program being summarily killed by the operating
469 system.</para>
470
471 <para>The maximum heap size also affects other garbage
472 collection parameters: when the amount of live data in the
473 heap exceeds a certain fraction of the maximum heap size,
474 compacting collection will be automatically enabled for the
475 oldest generation, and the <option>-F</option> parameter
476 will be reduced in order to avoid exceeding the maximum heap
477 size.</para>
478 </listitem>
479 </varlistentry>
480
481 <varlistentry>
482 <term>
483 <option>-t</option><optional><replaceable>file</replaceable></optional>
484 <indexterm><primary><option>-t</option></primary><secondary>RTS option</secondary></indexterm>
485 </term>
486 <term>
487 <option>-s</option><optional><replaceable>file</replaceable></optional>
488 <indexterm><primary><option>-s</option></primary><secondary>RTS option</secondary></indexterm>
489 </term>
490 <term>
491 <option>-S</option><optional><replaceable>file</replaceable></optional>
492 <indexterm><primary><option>-S</option></primary><secondary>RTS option</secondary></indexterm>
493 </term>
494 <term>
495 <option>--machine-readable</option>
496 <indexterm><primary><option>--machine-readable</option></primary><secondary>RTS option</secondary></indexterm>
497 </term>
498 <listitem>
499 <para>These options produce runtime-system statistics, such
500 as the amount of time spent executing the program and in the
501 garbage collector, the amount of memory allocated, the
502 maximum size of the heap, and so on. The three
503 variants give different levels of detail:
504 <option>-t</option> produces a single line of output in the
505 same format as GHC's <option>-Rghc-timing</option> option,
506 <option>-s</option> produces a more detailed summary at the
507 end of the program, and <option>-S</option> additionally
508 produces information about each and every garbage
509 collection.</para>
510
511 <para>The output is placed in
512 <replaceable>file</replaceable>. If
513 <replaceable>file</replaceable> is omitted, then the output
514 is sent to <constant>stderr</constant>.</para>
515
516 <para>
517 If you use the <literal>-t</literal> flag then, when your
518 program finishes, you will see something like this:
519 </para>
520
521 <programlisting>
522 &lt;&lt;ghc: 36169392 bytes, 69 GCs, 603392/1065272 avg/max bytes residency (2 samples), 3M in use, 0.00 INIT (0.00 elapsed), 0.02 MUT (0.02 elapsed), 0.07 GC (0.07 elapsed) :ghc&gt;&gt;
523 </programlisting>
524
525 <para>
526 This tells you:
527 </para>
528
529 <itemizedlist>
530 <listitem>
531 <para>
532 The total bytes allocated by the program. This may be less
533 than the peak memory use, as some may be freed.
534 </para>
535 </listitem>
536 <listitem>
537 <para>
538 The total number of garbage collections that occurred.
539 </para>
540 </listitem>
541 <listitem>
542 <para>
543 The average and maximum space used by your program.
544 This is only checked during major garbage collections, so it
545 is only an approximation; the number of samples tells you how
546 many times it is checked.
547 </para>
548 </listitem>
549 <listitem>
550 <para>
551 The peak memory the RTS has allocated from the OS.
552 </para>
553 </listitem>
554 <listitem>
555 <para>
556 The amount of CPU time and elapsed wall clock time while
557 initialising the runtime system (INIT), running the program
558 itself (MUT, the mutator), and garbage collecting (GC).
559 </para>
560 </listitem>
561 </itemizedlist>
562
563 <para>
564 You can also get this in a more future-proof, machine readable
565 format, with <literal>-t --machine-readable</literal>:
566 </para>
567
568 <programlisting>
569 [("bytes allocated", "36169392")
570 ,("num_GCs", "69")
571 ,("average_bytes_used", "603392")
572 ,("max_bytes_used", "1065272")
573 ,("num_byte_usage_samples", "2")
574 ,("peak_megabytes_allocated", "3")
575 ,("init_cpu_seconds", "0.00")
576 ,("init_wall_seconds", "0.00")
577 ,("mutator_cpu_seconds", "0.02")
578 ,("mutator_wall_seconds", "0.02")
579 ,("GC_cpu_seconds", "0.07")
580 ,("GC_wall_seconds", "0.07")
581 ]
582 </programlisting>
583
584 <para>
585 If you use the <literal>-s</literal> flag then, when your
586 program finishes, you will see something like this (the exact
587 details will vary depending on what sort of RTS you have, e.g.
588 you will only see profiling data if your RTS is compiled for
589 profiling):
590 </para>
591
592 <programlisting>
593 36,169,392 bytes allocated in the heap
594 4,057,632 bytes copied during GC
595 1,065,272 bytes maximum residency (2 sample(s))
596 54,312 bytes maximum slop
597 3 MB total memory in use (0 MB lost due to fragmentation)
598
599 Generation 0: 67 collections, 0 parallel, 0.04s, 0.03s elapsed
600 Generation 1: 2 collections, 0 parallel, 0.03s, 0.04s elapsed
601
602 SPARKS: 359207 (557 converted, 149591 pruned)
603
604 INIT time 0.00s ( 0.00s elapsed)
605 MUT time 0.01s ( 0.02s elapsed)
606 GC time 0.07s ( 0.07s elapsed)
607 EXIT time 0.00s ( 0.00s elapsed)
608 Total time 0.08s ( 0.09s elapsed)
609
610 %GC time 89.5% (75.3% elapsed)
611
612 Alloc rate 4,520,608,923 bytes per MUT second
613
614 Productivity 10.5% of total user, 9.1% of total elapsed
615 </programlisting>
616
617 <itemizedlist>
618 <listitem>
619 <para>
620 The "bytes allocated in the heap" is the total bytes allocated
621 by the program. This may be less than the peak memory use, as
622 some may be freed.
623 </para>
624 </listitem>
625 <listitem>
626 <para>
627 GHC uses a copying garbage collector. "bytes copied during GC"
628 tells you how many bytes it had to copy during garbage collection.
629 </para>
630 </listitem>
631 <listitem>
632 <para>
633 The maximum space actually used by your program is the
634 "bytes maximum residency" figure. This is only checked during
635 major garbage collections, so it is only an approximation;
636 the number of samples tells you how many times it is checked.
637 </para>
638 </listitem>
639 <listitem>
640 <para>
641 The "bytes maximum slop" tells you the most space that is ever
642 wasted due to the way GHC packs data into so-called "megablocks".
643 </para>
644 </listitem>
645 <listitem>
646 <para>
647 The "total memory in use" tells you the peak memory the RTS has
648 allocated from the OS.
649 </para>
650 </listitem>
651 <listitem>
652 <para>
653 Next there is information about the garbage collections done.
654 For each generation it says how many garbage collections were
655 done, how many of those collections used multiple threads,
656 the total CPU time used for garbage collecting that generation,
657 and the total wall clock time elapsed while garbage collecting
658 that generation.
659 </para>
660 </listitem>
661 <listitem>
662 <para>The <literal>SPARKS</literal> statistic refers to the
663 use of <literal>Control.Parallel.par</literal> and related
664 functionality in the program. Each spark represents a call
665 to <literal>par</literal>; a spark is "converted" when it is
666 executed in parallel; and a spark is "pruned" when it is
667 found to be already evaluated and is discarded from the pool
668 by the garbage collector. Any remaining sparks are
669 discarded at the end of execution, so "converted" plus
670 "pruned" does not necessarily add up to the total.</para>
671 </listitem>
672 <listitem>
673 <para>
674 Next there is the CPU time and wall clock time elapsedm broken
675 down by what the runtiem system was doing at the time.
676 INIT is the runtime system initialisation.
677 MUT is the mutator time, i.e. the time spent actually running
678 your code.
679 GC is the time spent doing garbage collection.
680 RP is the time spent doing retainer profiling.
681 PROF is the time spent doing other profiling.
682 EXIT is the runtime system shutdown time.
683 And finally, Total is, of course, the total.
684 </para>
685 <para>
686 %GC time tells you what percentage GC is of Total.
687 "Alloc rate" tells you the "bytes allocated in the heap" divided
688 by the MUT CPU time.
689 "Productivity" tells you what percentage of the Total CPU and wall
690 clock elapsed times are spent in the mutator (MUT).
691 </para>
692 </listitem>
693 </itemizedlist>
694
695 <para>
696 The <literal>-S</literal> flag, as well as giving the same
697 output as the <literal>-s</literal> flag, prints information
698 about each GC as it happens:
699 </para>
700
701 <programlisting>
702 Alloc Copied Live GC GC TOT TOT Page Flts
703 bytes bytes bytes user elap user elap
704 528496 47728 141512 0.01 0.02 0.02 0.02 0 0 (Gen: 1)
705 [...]
706 524944 175944 1726384 0.00 0.00 0.08 0.11 0 0 (Gen: 0)
707 </programlisting>
708
709 <para>
710 For each garbage collection, we print:
711 </para>
712
713 <itemizedlist>
714 <listitem>
715 <para>
716 How many bytes we allocated this garbage collection.
717 </para>
718 </listitem>
719 <listitem>
720 <para>
721 How many bytes we copied this garbage collection.
722 </para>
723 </listitem>
724 <listitem>
725 <para>
726 How many bytes are currently live.
727 </para>
728 </listitem>
729 <listitem>
730 <para>
731 How long this garbage collection took (CPU time and elapsed
732 wall clock time).
733 </para>
734 </listitem>
735 <listitem>
736 <para>
737 How long the program has been running (CPU time and elapsed
738 wall clock time).
739 </para>
740 </listitem>
741 <listitem>
742 <para>
743 How many page faults occured this garbage collection.
744 </para>
745 </listitem>
746 <listitem>
747 <para>
748 How many page faults occured since the end of the last garbage
749 collection.
750 </para>
751 </listitem>
752 <listitem>
753 <para>
754 Which generation is being garbage collected.
755 </para>
756 </listitem>
757 </itemizedlist>
758
759 </listitem>
760 </varlistentry>
761 </variablelist>
762
763 </sect2>
764
765 <sect2>
766 <title>RTS options for concurrency and parallelism</title>
767
768 <para>The RTS options related to concurrency are described in
769 <xref linkend="using-concurrent" />, and those for parallelism in
770 <xref linkend="parallel-options"/>.</para>
771 </sect2>
772
773 <sect2 id="rts-profiling">
774 <title>RTS options for profiling</title>
775
776 <para>Most profiling runtime options are only available when you
777 compile your program for profiling (see
778 <xref linkend="prof-compiler-options" />, and
779 <xref linkend="rts-options-heap-prof" /> for the runtime options).
780 However, there is one profiling option that is available
781 for ordinary non-profiled executables:</para>
782
783 <variablelist>
784 <varlistentry>
785 <term>
786 <option>-hT</option>
787 <indexterm><primary><option>-hT</option></primary><secondary>RTS
788 option</secondary></indexterm>
789 </term>
790 <listitem>
791 <para>Generates a basic heap profile, in the
792 file <literal><replaceable>prog</replaceable>.hp</literal>.
793 To produce the heap profile graph,
794 use <command>hp2ps</command> (see <xref linkend="hp2ps"
795 />). The basic heap profile is broken down by data
796 constructor, with other types of closures (functions, thunks,
797 etc.) grouped into broad categories
798 (e.g. <literal>FUN</literal>, <literal>THUNK</literal>). To
799 get a more detailed profile, use the full profiling
800 support (<xref linkend="profiling" />).</para>
801 </listitem>
802 </varlistentry>
803 </variablelist>
804 </sect2>
805
806 <sect2 id="rts-options-debugging">
807 <title>RTS options for hackers, debuggers, and over-interested
808 souls</title>
809
810 <indexterm><primary>RTS options, hacking/debugging</primary></indexterm>
811
812 <para>These RTS options might be used (a)&nbsp;to avoid a GHC bug,
813 (b)&nbsp;to see &ldquo;what's really happening&rdquo;, or
814 (c)&nbsp;because you feel like it. Not recommended for everyday
815 use!</para>
816
817 <variablelist>
818
819 <varlistentry>
820 <term>
821 <option>-B</option>
822 <indexterm><primary><option>-B</option></primary><secondary>RTS option</secondary></indexterm>
823 </term>
824 <listitem>
825 <para>Sound the bell at the start of each (major) garbage
826 collection.</para>
827
828 <para>Oddly enough, people really do use this option! Our
829 pal in Durham (England), Paul Callaghan, writes: &ldquo;Some
830 people here use it for a variety of
831 purposes&mdash;honestly!&mdash;e.g., confirmation that the
832 code/machine is doing something, infinite loop detection,
833 gauging cost of recently added code. Certain people can even
834 tell what stage &lsqb;the program&rsqb; is in by the beep
835 pattern. But the major use is for annoying others in the
836 same office&hellip;&rdquo;</para>
837 </listitem>
838 </varlistentry>
839
840 <varlistentry>
841 <term>
842 <option>-D</option><replaceable>num</replaceable>
843 <indexterm><primary>-D</primary><secondary>RTS option</secondary></indexterm>
844 </term>
845 <listitem>
846 <para>An RTS debugging flag; varying quantities of output
847 depending on which bits are set in
848 <replaceable>num</replaceable>. Only works if the RTS was
849 compiled with the <option>DEBUG</option> option.</para>
850 </listitem>
851 </varlistentry>
852
853 <varlistentry>
854 <term>
855 <option>-r</option><replaceable>file</replaceable>
856 <indexterm><primary><option>-r</option></primary><secondary>RTS option</secondary></indexterm>
857 <indexterm><primary>ticky ticky profiling</primary></indexterm>
858 <indexterm><primary>profiling</primary><secondary>ticky ticky</secondary></indexterm>
859 </term>
860 <listitem>
861 <para>Produce &ldquo;ticky-ticky&rdquo; statistics at the
862 end of the program run. The <replaceable>file</replaceable>
863 business works just like on the <option>-S</option> RTS
864 option (above).</para>
865
866 <para>&ldquo;Ticky-ticky&rdquo; statistics are counts of
867 various program actions (updates, enters, etc.) The program
868 must have been compiled using
869 <option>-ticky</option><indexterm><primary><option>-ticky</option></primary></indexterm>
870 (a.k.a. &ldquo;ticky-ticky profiling&rdquo;), and, for it to
871 be really useful, linked with suitable system libraries.
872 Not a trivial undertaking: consult the installation guide on
873 how to set things up for easy &ldquo;ticky-ticky&rdquo;
874 profiling. For more information, see <xref
875 linkend="ticky-ticky"/>.</para>
876 </listitem>
877 </varlistentry>
878
879 <varlistentry>
880 <term>
881 <option>-xc</option>
882 <indexterm><primary><option>-xc</option></primary><secondary>RTS option</secondary></indexterm>
883 </term>
884 <listitem>
885 <para>(Only available when the program is compiled for
886 profiling.) When an exception is raised in the program,
887 this option causes the current cost-centre-stack to be
888 dumped to <literal>stderr</literal>.</para>
889
890 <para>This can be particularly useful for debugging: if your
891 program is complaining about a <literal>head []</literal>
892 error and you haven't got a clue which bit of code is
893 causing it, compiling with <literal>-prof
894 -auto-all</literal> and running with <literal>+RTS -xc
895 -RTS</literal> will tell you exactly the call stack at the
896 point the error was raised.</para>
897
898 <para>The output contains one line for each exception raised
899 in the program (the program might raise and catch several
900 exceptions during its execution), where each line is of the
901 form:</para>
902
903 <screen>
904 &lt; cc<subscript>1</subscript>, ..., cc<subscript>n</subscript> &gt;
905 </screen>
906 <para>each <literal>cc</literal><subscript>i</subscript> is
907 a cost centre in the program (see <xref
908 linkend="cost-centres"/>), and the sequence represents the
909 &ldquo;call stack&rdquo; at the point the exception was
910 raised. The leftmost item is the innermost function in the
911 call stack, and the rightmost item is the outermost
912 function.</para>
913
914 </listitem>
915 </varlistentry>
916
917 <varlistentry>
918 <term>
919 <option>-Z</option>
920 <indexterm><primary><option>-Z</option></primary><secondary>RTS option</secondary></indexterm>
921 </term>
922 <listitem>
923 <para>Turn <emphasis>off</emphasis> &ldquo;update-frame
924 squeezing&rdquo; at garbage-collection time. (There's no
925 particularly good reason to turn it off, except to ensure
926 the accuracy of certain data collected regarding thunk entry
927 counts.)</para>
928 </listitem>
929 </varlistentry>
930 </variablelist>
931
932 </sect2>
933
934 <sect2 id="rts-hooks">
935 <title>&ldquo;Hooks&rdquo; to change RTS behaviour</title>
936
937 <indexterm><primary>hooks</primary><secondary>RTS</secondary></indexterm>
938 <indexterm><primary>RTS hooks</primary></indexterm>
939 <indexterm><primary>RTS behaviour, changing</primary></indexterm>
940
941 <para>GHC lets you exercise rudimentary control over the RTS
942 settings for any given program, by compiling in a
943 &ldquo;hook&rdquo; that is called by the run-time system. The RTS
944 contains stub definitions for all these hooks, but by writing your
945 own version and linking it on the GHC command line, you can
946 override the defaults.</para>
947
948 <para>Owing to the vagaries of DLL linking, these hooks don't work
949 under Windows when the program is built dynamically.</para>
950
951 <para>The hook <literal>ghc_rts_opts</literal><indexterm><primary><literal>ghc_rts_opts</literal></primary>
952 </indexterm>lets you set RTS
953 options permanently for a given program. A common use for this is
954 to give your program a default heap and/or stack size that is
955 greater than the default. For example, to set <literal>-H128m
956 -K1m</literal>, place the following definition in a C source
957 file:</para>
958
959 <programlisting>
960 char *ghc_rts_opts = "-H128m -K1m";
961 </programlisting>
962
963 <para>Compile the C file, and include the object file on the
964 command line when you link your Haskell program.</para>
965
966 <para>These flags are interpreted first, before any RTS flags from
967 the <literal>GHCRTS</literal> environment variable and any flags
968 on the command line.</para>
969
970 <para>You can also change the messages printed when the runtime
971 system &ldquo;blows up,&rdquo; e.g., on stack overflow. The hooks
972 for these are as follows:</para>
973
974 <variablelist>
975
976 <varlistentry>
977 <term>
978 <function>void OutOfHeapHook (unsigned long, unsigned long)</function>
979 <indexterm><primary><function>OutOfHeapHook</function></primary></indexterm>
980 </term>
981 <listitem>
982 <para>The heap-overflow message.</para>
983 </listitem>
984 </varlistentry>
985
986 <varlistentry>
987 <term>
988 <function>void StackOverflowHook (long int)</function>
989 <indexterm><primary><function>StackOverflowHook</function></primary></indexterm>
990 </term>
991 <listitem>
992 <para>The stack-overflow message.</para>
993 </listitem>
994 </varlistentry>
995
996 <varlistentry>
997 <term>
998 <function>void MallocFailHook (long int)</function>
999 <indexterm><primary><function>MallocFailHook</function></primary></indexterm>
1000 </term>
1001 <listitem>
1002 <para>The message printed if <function>malloc</function>
1003 fails.</para>
1004 </listitem>
1005 </varlistentry>
1006 </variablelist>
1007
1008 <para>For examples of the use of these hooks, see GHC's own
1009 versions in the file
1010 <filename>ghc/compiler/parser/hschooks.c</filename> in a GHC
1011 source tree.</para>
1012 </sect2>
1013
1014 <sect2>
1015 <title>Getting information about the RTS</title>
1016
1017 <indexterm><primary>RTS</primary></indexterm>
1018
1019 <para>It is possible to ask the RTS to give some information about
1020 itself. To do this, use the <option>--info</option> flag, e.g.</para>
1021 <screen>
1022 $ ./a.out +RTS --info
1023 [("GHC RTS", "Yes")
1024 ,("GHC version", "6.7")
1025 ,("RTS way", "rts_p")
1026 ,("Host platform", "x86_64-unknown-linux")
1027 ,("Build platform", "x86_64-unknown-linux")
1028 ,("Target platform", "x86_64-unknown-linux")
1029 ,("Compiler unregisterised", "NO")
1030 ,("Tables next to code", "YES")
1031 ]
1032 </screen>
1033 <para>The information is formatted such that it can be read as a
1034 of type <literal>[(String, String)]</literal>.</para>
1035 </sect2>
1036 </sect1>
1037
1038 <!-- Emacs stuff:
1039 ;;; Local Variables: ***
1040 ;;; mode: xml ***
1041 ;;; sgml-parent-document: ("users_guide.xml" "book" "chapter" "sect1") ***
1042 ;;; End: ***
1043 -->