Commit | Line | Data |
---|---|---|
07d83abe MV |
1 | @c -*-texinfo-*- |
2 | @c This is part of the GNU Guile Reference Manual. | |
d3d66147 | 3 | @c Copyright (C) 1996, 1997, 2000, 2001, 2002, 2003, 2004, 2007, 2009 |
07d83abe MV |
4 | @c Free Software Foundation, Inc. |
5 | @c See the file guile.texi for copying conditions. | |
6 | ||
7 | @page | |
8 | @node Scheduling | |
9 | @section Threads, Mutexes, Asyncs and Dynamic Roots | |
10 | ||
07d83abe MV |
11 | @menu |
12 | * Arbiters:: Synchronization primitives. | |
13 | * Asyncs:: Asynchronous procedure invocation. | |
07d83abe | 14 | * Threads:: Multiple threads of execution. |
2567692a | 15 | * Mutexes and Condition Variables:: Synchronization primitives. |
b4fddbbe | 16 | * Blocking:: How to block properly in guile mode. |
2567692a | 17 | * Critical Sections:: Avoiding concurrency and reentries. |
b4fddbbe | 18 | * Fluids and Dynamic States:: Thread-local variables, etc. |
07d83abe MV |
19 | * Parallel Forms:: Parallel execution of forms. |
20 | @end menu | |
21 | ||
22 | ||
23 | @node Arbiters | |
24 | @subsection Arbiters | |
07d83abe MV |
25 | @cindex arbiters |
26 | ||
e136aab0 KR |
27 | Arbiters are synchronization objects, they can be used by threads to |
28 | control access to a shared resource. An arbiter can be locked to | |
29 | indicate a resource is in use, and unlocked when done. | |
07d83abe | 30 | |
b4fddbbe MV |
31 | An arbiter is like a light-weight mutex (@pxref{Mutexes and Condition |
32 | Variables}). It uses less memory and may be faster, but there's no | |
33 | way for a thread to block waiting on an arbiter, it can only test and | |
34 | get the status returned. | |
07d83abe MV |
35 | |
36 | @deffn {Scheme Procedure} make-arbiter name | |
37 | @deffnx {C Function} scm_make_arbiter (name) | |
cdf1ad3b MV |
38 | Return an object of type arbiter and name @var{name}. Its |
39 | state is initially unlocked. Arbiters are a way to achieve | |
40 | process synchronization. | |
07d83abe MV |
41 | @end deffn |
42 | ||
43 | @deffn {Scheme Procedure} try-arbiter arb | |
44 | @deffnx {C Function} scm_try_arbiter (arb) | |
cdf1ad3b MV |
45 | If @var{arb} is unlocked, then lock it and return @code{#t}. |
46 | If @var{arb} is already locked, then do nothing and return | |
47 | @code{#f}. | |
07d83abe MV |
48 | @end deffn |
49 | ||
50 | @deffn {Scheme Procedure} release-arbiter arb | |
51 | @deffnx {C Function} scm_release_arbiter (arb) | |
e136aab0 KR |
52 | If @var{arb} is locked, then unlock it and return @code{#t}. If |
53 | @var{arb} is already unlocked, then do nothing and return @code{#f}. | |
54 | ||
55 | Typical usage is for the thread which locked an arbiter to later | |
56 | release it, but that's not required, any thread can release it. | |
07d83abe MV |
57 | @end deffn |
58 | ||
59 | ||
60 | @node Asyncs | |
61 | @subsection Asyncs | |
62 | ||
63 | @cindex asyncs | |
64 | @cindex user asyncs | |
65 | @cindex system asyncs | |
66 | ||
1021bb7a | 67 | Asyncs are a means of deferring the execution of Scheme code until it is |
07d83abe MV |
68 | safe to do so. |
69 | ||
70 | Guile provides two kinds of asyncs that share the basic concept but are | |
71 | otherwise quite different: system asyncs and user asyncs. System asyncs | |
72 | are integrated into the core of Guile and are executed automatically | |
73 | when the system is in a state to allow the execution of Scheme code. | |
74 | For example, it is not possible to execute Scheme code in a POSIX signal | |
75 | handler, but such a signal handler can queue a system async to be | |
76 | executed in the near future, when it is safe to do so. | |
77 | ||
78 | System asyncs can also be queued for threads other than the current one. | |
79 | This way, you can cause threads to asynchronously execute arbitrary | |
80 | code. | |
81 | ||
82 | User asyncs offer a convenient means of queueing procedures for future | |
83 | execution and triggering this execution. They will not be executed | |
84 | automatically. | |
85 | ||
86 | @menu | |
74926120 NJ |
87 | * System asyncs:: |
88 | * User asyncs:: | |
07d83abe MV |
89 | @end menu |
90 | ||
91 | @node System asyncs | |
92 | @subsubsection System asyncs | |
93 | ||
94 | To cause the future asynchronous execution of a procedure in a given | |
95 | thread, use @code{system-async-mark}. | |
96 | ||
97 | Automatic invocation of system asyncs can be temporarily disabled by | |
98 | calling @code{call-with-blocked-asyncs}. This function works by | |
99 | temporarily increasing the @emph{async blocking level} of the current | |
100 | thread while a given procedure is running. The blocking level starts | |
101 | out at zero, and whenever a safe point is reached, a blocking level | |
102 | greater than zero will prevent the execution of queued asyncs. | |
103 | ||
104 | Analogously, the procedure @code{call-with-unblocked-asyncs} will | |
105 | temporarily decrease the blocking level of the current thread. You | |
106 | can use it when you want to disable asyncs by default and only allow | |
107 | them temporarily. | |
108 | ||
109 | In addition to the C versions of @code{call-with-blocked-asyncs} and | |
110 | @code{call-with-unblocked-asyncs}, C code can use | |
661ae7ab MV |
111 | @code{scm_dynwind_block_asyncs} and @code{scm_dynwind_unblock_asyncs} |
112 | inside a @dfn{dynamic context} (@pxref{Dynamic Wind}) to block or | |
113 | unblock system asyncs temporarily. | |
07d83abe MV |
114 | |
115 | @deffn {Scheme Procedure} system-async-mark proc [thread] | |
116 | @deffnx {C Function} scm_system_async_mark (proc) | |
117 | @deffnx {C Function} scm_system_async_mark_for_thread (proc, thread) | |
118 | Mark @var{proc} (a procedure with zero arguments) for future execution | |
119 | in @var{thread}. When @var{proc} has already been marked for | |
120 | @var{thread} but has not been executed yet, this call has no effect. | |
121 | When @var{thread} is omitted, the thread that called | |
122 | @code{system-async-mark} is used. | |
123 | ||
124 | This procedure is not safe to be called from signal handlers. Use | |
125 | @code{scm_sigaction} or @code{scm_sigaction_for_thread} to install | |
126 | signal handlers. | |
127 | @end deffn | |
128 | ||
07d83abe MV |
129 | @deffn {Scheme Procedure} call-with-blocked-asyncs proc |
130 | @deffnx {C Function} scm_call_with_blocked_asyncs (proc) | |
07d83abe MV |
131 | Call @var{proc} and block the execution of system asyncs by one level |
132 | for the current thread while it is running. Return the value returned | |
133 | by @var{proc}. For the first two variants, call @var{proc} with no | |
134 | arguments; for the third, call it with @var{data}. | |
135 | @end deffn | |
136 | ||
1021bb7a NJ |
137 | @deftypefn {C Function} {void *} scm_c_call_with_blocked_asyncs (void * (*proc) (void *data), void *data) |
138 | The same but with a C function @var{proc} instead of a Scheme thunk. | |
139 | @end deftypefn | |
140 | ||
07d83abe MV |
141 | @deffn {Scheme Procedure} call-with-unblocked-asyncs proc |
142 | @deffnx {C Function} scm_call_with_unblocked_asyncs (proc) | |
07d83abe MV |
143 | Call @var{proc} and unblock the execution of system asyncs by one |
144 | level for the current thread while it is running. Return the value | |
145 | returned by @var{proc}. For the first two variants, call @var{proc} | |
146 | with no arguments; for the third, call it with @var{data}. | |
147 | @end deffn | |
148 | ||
1021bb7a NJ |
149 | @deftypefn {C Function} {void *} scm_c_call_with_unblocked_asyncs (void *(*proc) (void *data), void *data) |
150 | The same but with a C function @var{proc} instead of a Scheme thunk. | |
151 | @end deftypefn | |
152 | ||
661ae7ab | 153 | @deftypefn {C Function} void scm_dynwind_block_asyncs () |
1021bb7a NJ |
154 | During the current dynwind context, increase the blocking of asyncs by |
155 | one level. This function must be used inside a pair of calls to | |
661ae7ab | 156 | @code{scm_dynwind_begin} and @code{scm_dynwind_end} (@pxref{Dynamic |
1021bb7a | 157 | Wind}). |
07d83abe MV |
158 | @end deftypefn |
159 | ||
661ae7ab | 160 | @deftypefn {C Function} void scm_dynwind_unblock_asyncs () |
1021bb7a NJ |
161 | During the current dynwind context, decrease the blocking of asyncs by |
162 | one level. This function must be used inside a pair of calls to | |
661ae7ab | 163 | @code{scm_dynwind_begin} and @code{scm_dynwind_end} (@pxref{Dynamic |
1021bb7a | 164 | Wind}). |
07d83abe MV |
165 | @end deftypefn |
166 | ||
167 | @node User asyncs | |
168 | @subsubsection User asyncs | |
169 | ||
170 | A user async is a pair of a thunk (a parameterless procedure) and a | |
171 | mark. Setting the mark on a user async will cause the thunk to be | |
172 | executed when the user async is passed to @code{run-asyncs}. Setting | |
173 | the mark more than once is satisfied by one execution of the thunk. | |
174 | ||
175 | User asyncs are created with @code{async}. They are marked with | |
176 | @code{async-mark}. | |
177 | ||
178 | @deffn {Scheme Procedure} async thunk | |
179 | @deffnx {C Function} scm_async (thunk) | |
180 | Create a new user async for the procedure @var{thunk}. | |
181 | @end deffn | |
182 | ||
183 | @deffn {Scheme Procedure} async-mark a | |
184 | @deffnx {C Function} scm_async_mark (a) | |
185 | Mark the user async @var{a} for future execution. | |
186 | @end deffn | |
187 | ||
188 | @deffn {Scheme Procedure} run-asyncs list_of_a | |
189 | @deffnx {C Function} scm_run_asyncs (list_of_a) | |
190 | Execute all thunks from the marked asyncs of the list @var{list_of_a}. | |
191 | @end deffn | |
192 | ||
07d83abe MV |
193 | @node Threads |
194 | @subsection Threads | |
195 | @cindex threads | |
196 | @cindex Guile threads | |
197 | @cindex POSIX threads | |
198 | ||
cdf1ad3b MV |
199 | @deffn {Scheme Procedure} all-threads |
200 | @deffnx {C Function} scm_all_threads () | |
201 | Return a list of all threads. | |
202 | @end deffn | |
203 | ||
204 | @deffn {Scheme Procedure} current-thread | |
205 | @deffnx {C Function} scm_current_thread () | |
206 | Return the thread that called this function. | |
207 | @end deffn | |
07d83abe MV |
208 | |
209 | @c begin (texi-doc-string "guile" "call-with-new-thread") | |
23f2b9a3 | 210 | @deffn {Scheme Procedure} call-with-new-thread thunk [handler] |
b4fddbbe MV |
211 | Call @code{thunk} in a new thread and with a new dynamic state, |
212 | returning the new thread. The procedure @var{thunk} is called via | |
213 | @code{with-continuation-barrier}. | |
07d83abe | 214 | |
b4fddbbe MV |
215 | When @var{handler} is specified, then @var{thunk} is called from |
216 | within a @code{catch} with tag @code{#t} that has @var{handler} as its | |
217 | handler. This catch is established inside the continuation barrier. | |
07d83abe | 218 | |
b4fddbbe MV |
219 | Once @var{thunk} or @var{handler} returns, the return value is made |
220 | the @emph{exit value} of the thread and the thread is terminated. | |
07d83abe MV |
221 | @end deffn |
222 | ||
b4fddbbe MV |
223 | @deftypefn {C Function} SCM scm_spawn_thread (scm_t_catch_body body, void *body_data, scm_t_catch_handler handler, void *handler_data) |
224 | Call @var{body} in a new thread, passing it @var{body_data}, returning | |
225 | the new thread. The function @var{body} is called via | |
226 | @code{scm_c_with_continuation_barrier}. | |
227 | ||
228 | When @var{handler} is non-@code{NULL}, @var{body} is called via | |
229 | @code{scm_internal_catch} with tag @code{SCM_BOOL_T} that has | |
230 | @var{handler} and @var{handler_data} as the handler and its data. This | |
231 | catch is established inside the continuation barrier. | |
232 | ||
233 | Once @var{body} or @var{handler} returns, the return value is made the | |
234 | @emph{exit value} of the thread and the thread is terminated. | |
235 | @end deftypefn | |
236 | ||
6180e336 NJ |
237 | @deffn {Scheme Procedure} thread? obj |
238 | @deffnx {C Function} scm_thread_p (obj) | |
239 | Return @code{#t} iff @var{obj} is a thread; otherwise, return | |
240 | @code{#f}. | |
241 | @end deffn | |
242 | ||
07d83abe | 243 | @c begin (texi-doc-string "guile" "join-thread") |
6180e336 | 244 | @deffn {Scheme Procedure} join-thread thread [timeout [timeoutval]] |
300b1ae5 | 245 | @deffnx {C Function} scm_join_thread (thread) |
6180e336 | 246 | @deffnx {C Function} scm_join_thread_timed (thread, timeout, timeoutval) |
b4fddbbe MV |
247 | Wait for @var{thread} to terminate and return its exit value. Threads |
248 | that have not been created with @code{call-with-new-thread} or | |
74926120 | 249 | @code{scm_spawn_thread} have an exit value of @code{#f}. When |
6180e336 | 250 | @var{timeout} is given, it specifies a point in time where the waiting |
74926120 NJ |
251 | should be aborted. It can be either an integer as returned by |
252 | @code{current-time} or a pair as returned by @code{gettimeofday}. | |
253 | When the waiting is aborted, @var{timeoutval} is returned (if it is | |
6180e336 | 254 | specified; @code{#f} is returned otherwise). |
07d83abe MV |
255 | @end deffn |
256 | ||
cdf1ad3b MV |
257 | @deffn {Scheme Procedure} thread-exited? thread |
258 | @deffnx {C Function} scm_thread_exited_p (thread) | |
259 | Return @code{#t} iff @var{thread} has exited. | |
260 | @end deffn | |
261 | ||
07d83abe MV |
262 | @c begin (texi-doc-string "guile" "yield") |
263 | @deffn {Scheme Procedure} yield | |
264 | If one or more threads are waiting to execute, calling yield forces an | |
265 | immediate context switch to one of them. Otherwise, yield has no effect. | |
266 | @end deffn | |
267 | ||
07e02175 LC |
268 | @deffn {Scheme Procedure} cancel-thread thread |
269 | @deffnx {C Function} scm_cancel_thread (thread) | |
270 | Asynchronously notify @var{thread} to exit. Immediately after | |
271 | receiving this notification, @var{thread} will call its cleanup handler | |
272 | (if one has been set) and then terminate, aborting any evaluation that | |
273 | is in progress. | |
274 | ||
275 | Because Guile threads are isomorphic with POSIX threads, @var{thread} | |
276 | will not receive its cancellation signal until it reaches a cancellation | |
277 | point. See your operating system's POSIX threading documentation for | |
278 | more information on cancellation points; note that in Guile, unlike | |
279 | native POSIX threads, a thread can receive a cancellation notification | |
280 | while attempting to lock a mutex. | |
281 | @end deffn | |
282 | ||
283 | @deffn {Scheme Procedure} set-thread-cleanup! thread proc | |
284 | @deffnx {C Function} scm_set_thread_cleanup_x (thread, proc) | |
285 | Set @var{proc} as the cleanup handler for the thread @var{thread}. | |
286 | @var{proc}, which must be a thunk, will be called when @var{thread} | |
287 | exits, either normally or by being canceled. Thread cleanup handlers | |
288 | can be used to perform useful tasks like releasing resources, such as | |
289 | locked mutexes, when thread exit cannot be predicted. | |
290 | ||
291 | The return value of @var{proc} will be set as the @emph{exit value} of | |
292 | @var{thread}. | |
293 | ||
294 | To remove a cleanup handler, pass @code{#f} for @var{proc}. | |
295 | @end deffn | |
296 | ||
297 | @deffn {Scheme Procedure} thread-cleanup thread | |
298 | @deffnx {C Function} scm_thread_cleanup (thread) | |
299 | Return the cleanup handler currently installed for the thread | |
300 | @var{thread}. If no cleanup handler is currently installed, | |
301 | thread-cleanup returns @code{#f}. | |
302 | @end deffn | |
303 | ||
07d83abe MV |
304 | Higher level thread procedures are available by loading the |
305 | @code{(ice-9 threads)} module. These provide standardized | |
3cf066df | 306 | thread creation. |
07d83abe MV |
307 | |
308 | @deffn macro make-thread proc [args@dots{}] | |
309 | Apply @var{proc} to @var{args} in a new thread formed by | |
310 | @code{call-with-new-thread} using a default error handler that display | |
b4fddbbe MV |
311 | the error to the current error port. The @var{args@dots{}} |
312 | expressions are evaluated in the new thread. | |
07d83abe MV |
313 | @end deffn |
314 | ||
315 | @deffn macro begin-thread first [rest@dots{}] | |
316 | Evaluate forms @var{first} and @var{rest} in a new thread formed by | |
317 | @code{call-with-new-thread} using a default error handler that display | |
318 | the error to the current error port. | |
319 | @end deffn | |
320 | ||
2567692a MV |
321 | @node Mutexes and Condition Variables |
322 | @subsection Mutexes and Condition Variables | |
323 | @cindex mutex | |
324 | @cindex condition variable | |
325 | ||
326 | A mutex is a thread synchronization object, it can be used by threads | |
327 | to control access to a shared resource. A mutex can be locked to | |
328 | indicate a resource is in use, and other threads can then block on the | |
329 | mutex to wait for the resource (or can just test and do something else | |
330 | if not available). ``Mutex'' is short for ``mutual exclusion''. | |
331 | ||
332 | There are two types of mutexes in Guile, ``standard'' and | |
333 | ``recursive''. They're created by @code{make-mutex} and | |
334 | @code{make-recursive-mutex} respectively, the operation functions are | |
335 | then common to both. | |
336 | ||
337 | Note that for both types of mutex there's no protection against a | |
338 | ``deadly embrace''. For instance if one thread has locked mutex A and | |
339 | is waiting on mutex B, but another thread owns B and is waiting on A, | |
340 | then an endless wait will occur (in the current implementation). | |
341 | Acquiring requisite mutexes in a fixed order (like always A before B) | |
342 | in all threads is one way to avoid such problems. | |
343 | ||
344 | @sp 1 | |
6180e336 | 345 | @deffn {Scheme Procedure} make-mutex . flags |
2567692a | 346 | @deffnx {C Function} scm_make_mutex () |
9c9b203b | 347 | @deffnx {C Function} scm_make_mutex_with_flags (SCM flags) |
74926120 | 348 | Return a new mutex. It is initially unlocked. If @var{flags} is |
6180e336 | 349 | specified, it must be a list of symbols specifying configuration flags |
74926120 | 350 | for the newly-created mutex. The supported flags are: |
6180e336 NJ |
351 | @table @code |
352 | @item unchecked-unlock | |
353 | Unless this flag is present, a call to `unlock-mutex' on the returned | |
354 | mutex when it is already unlocked will cause an error to be signalled. | |
355 | ||
356 | @item allow-external-unlock | |
357 | Allow the returned mutex to be unlocked by the calling thread even if | |
358 | it was originally locked by a different thread. | |
359 | ||
360 | @item recursive | |
361 | The returned mutex will be recursive. | |
362 | ||
363 | @end table | |
364 | @end deffn | |
365 | ||
366 | @deffn {Scheme Procedure} mutex? obj | |
367 | @deffnx {C Function} scm_mutex_p (obj) | |
74926120 | 368 | Return @code{#t} iff @var{obj} is a mutex; otherwise, return |
6180e336 | 369 | @code{#f}. |
2567692a MV |
370 | @end deffn |
371 | ||
372 | @deffn {Scheme Procedure} make-recursive-mutex | |
373 | @deffnx {C Function} scm_make_recursive_mutex () | |
6180e336 NJ |
374 | Create a new recursive mutex. It is initially unlocked. Calling this |
375 | function is equivalent to calling `make-mutex' and specifying the | |
376 | @code{recursive} flag. | |
2567692a MV |
377 | @end deffn |
378 | ||
adc085f1 | 379 | @deffn {Scheme Procedure} lock-mutex mutex [timeout [owner]] |
2567692a | 380 | @deffnx {C Function} scm_lock_mutex (mutex) |
adc085f1 | 381 | @deffnx {C Function} scm_lock_mutex_timed (mutex, timeout, owner) |
74926120 | 382 | Lock @var{mutex}. If the mutex is already locked, then block and |
adc085f1 | 383 | return only when @var{mutex} has been acquired. |
2567692a | 384 | |
74926120 NJ |
385 | When @var{timeout} is given, it specifies a point in time where the |
386 | waiting should be aborted. It can be either an integer as returned | |
387 | by @code{current-time} or a pair as returned by @code{gettimeofday}. | |
388 | When the waiting is aborted, @code{#f} is returned. | |
6180e336 | 389 | |
adc085f1 | 390 | When @var{owner} is given, it specifies an owner for @var{mutex} other |
74926120 | 391 | than the calling thread. @var{owner} may also be @code{#f}, |
adc085f1 JG |
392 | indicating that the mutex should be locked but left unowned. |
393 | ||
2567692a MV |
394 | For standard mutexes (@code{make-mutex}), and error is signalled if |
395 | the thread has itself already locked @var{mutex}. | |
396 | ||
397 | For a recursive mutex (@code{make-recursive-mutex}), if the thread has | |
398 | itself already locked @var{mutex}, then a further @code{lock-mutex} | |
399 | call increments the lock count. An additional @code{unlock-mutex} | |
400 | will be required to finally release. | |
401 | ||
6180e336 | 402 | If @var{mutex} was locked by a thread that exited before unlocking it, |
74926120 | 403 | the next attempt to lock @var{mutex} will succeed, but |
6180e336 NJ |
404 | @code{abandoned-mutex-error} will be signalled. |
405 | ||
2567692a MV |
406 | When a system async (@pxref{System asyncs}) is activated for a thread |
407 | blocked in @code{lock-mutex}, the wait is interrupted and the async is | |
408 | executed. When the async returns, the wait resumes. | |
409 | @end deffn | |
410 | ||
661ae7ab MV |
411 | @deftypefn {C Function} void scm_dynwind_lock_mutex (SCM mutex) |
412 | Arrange for @var{mutex} to be locked whenever the current dynwind | |
413 | context is entered and to be unlocked when it is exited. | |
2567692a | 414 | @end deftypefn |
74926120 | 415 | |
2567692a MV |
416 | @deffn {Scheme Procedure} try-mutex mx |
417 | @deffnx {C Function} scm_try_mutex (mx) | |
418 | Try to lock @var{mutex} as per @code{lock-mutex}. If @var{mutex} can | |
419 | be acquired immediately then this is done and the return is @code{#t}. | |
420 | If @var{mutex} is locked by some other thread then nothing is done and | |
421 | the return is @code{#f}. | |
422 | @end deffn | |
423 | ||
6180e336 | 424 | @deffn {Scheme Procedure} unlock-mutex mutex [condvar [timeout]] |
2567692a | 425 | @deffnx {C Function} scm_unlock_mutex (mutex) |
6180e336 NJ |
426 | @deffnx {C Function} scm_unlock_mutex_timed (mutex, condvar, timeout) |
427 | Unlock @var{mutex}. An error is signalled if @var{mutex} is not locked | |
74926120 | 428 | and was not created with the @code{unchecked-unlock} flag set, or if |
6180e336 NJ |
429 | @var{mutex} is locked by a thread other than the calling thread and was |
430 | not created with the @code{allow-external-unlock} flag set. | |
431 | ||
432 | If @var{condvar} is given, it specifies a condition variable upon | |
433 | which the calling thread will wait to be signalled before returning. | |
74926120 | 434 | (This behavior is very similar to that of |
6180e336 NJ |
435 | @code{wait-condition-variable}, except that the mutex is left in an |
436 | unlocked state when the function returns.) | |
437 | ||
74926120 NJ |
438 | When @var{timeout} is also given, it specifies a point in time where |
439 | the waiting should be aborted. It can be either an integer as | |
440 | returned by @code{current-time} or a pair as returned by | |
441 | @code{gettimeofday}. When the waiting is aborted, @code{#f} is | |
6180e336 | 442 | returned. Otherwise the function returns @code{#t}. |
2567692a MV |
443 | @end deffn |
444 | ||
adc085f1 JG |
445 | @deffn {Scheme Procedure} mutex-owner mutex |
446 | @deffnx {C Function} scm_mutex_owner (mutex) | |
74926120 | 447 | Return the current owner of @var{mutex}, in the form of a thread or |
adc085f1 JG |
448 | @code{#f} (indicating no owner). Note that a mutex may be unowned but |
449 | still locked. | |
450 | @end deffn | |
451 | ||
452 | @deffn {Scheme Procedure} mutex-level mutex | |
453 | @deffnx {C Function} scm_mutex_level (mutex) | |
454 | Return the current lock level of @var{mutex}. If @var{mutex} is | |
455 | currently unlocked, this value will be 0; otherwise, it will be the | |
456 | number of times @var{mutex} has been recursively locked by its current | |
457 | owner. | |
458 | @end deffn | |
459 | ||
460 | @deffn {Scheme Procedure} mutex-locked? mutex | |
461 | @deffnx {C Function} scm_mutex_locked_p (mutex) | |
462 | Return @code{#t} if @var{mutex} is locked, regardless of ownership; | |
463 | otherwise, return @code{#f}. | |
464 | @end deffn | |
465 | ||
2567692a MV |
466 | @deffn {Scheme Procedure} make-condition-variable |
467 | @deffnx {C Function} scm_make_condition_variable () | |
468 | Return a new condition variable. | |
469 | @end deffn | |
470 | ||
6180e336 NJ |
471 | @deffn {Scheme Procedure} condition-variable? obj |
472 | @deffnx {C Function} scm_condition_variable_p (obj) | |
74926120 | 473 | Return @code{#t} iff @var{obj} is a condition variable; otherwise, |
6180e336 NJ |
474 | return @code{#f}. |
475 | @end deffn | |
476 | ||
2567692a MV |
477 | @deffn {Scheme Procedure} wait-condition-variable condvar mutex [time] |
478 | @deffnx {C Function} scm_wait_condition_variable (condvar, mutex, time) | |
479 | Wait until @var{condvar} has been signalled. While waiting, | |
480 | @var{mutex} is atomically unlocked (as with @code{unlock-mutex}) and | |
481 | is locked again when this function returns. When @var{time} is given, | |
482 | it specifies a point in time where the waiting should be aborted. It | |
483 | can be either a integer as returned by @code{current-time} or a pair | |
484 | as returned by @code{gettimeofday}. When the waiting is aborted, | |
485 | @code{#f} is returned. When the condition variable has in fact been | |
486 | signalled, @code{#t} is returned. The mutex is re-locked in any case | |
487 | before @code{wait-condition-variable} returns. | |
488 | ||
489 | When a system async is activated for a thread that is blocked in a | |
490 | call to @code{wait-condition-variable}, the waiting is interrupted, | |
491 | the mutex is locked, and the async is executed. When the async | |
492 | returns, the mutex is unlocked again and the waiting is resumed. When | |
493 | the thread block while re-acquiring the mutex, execution of asyncs is | |
494 | blocked. | |
495 | @end deffn | |
496 | ||
497 | @deffn {Scheme Procedure} signal-condition-variable condvar | |
498 | @deffnx {C Function} scm_signal_condition_variable (condvar) | |
499 | Wake up one thread that is waiting for @var{condvar}. | |
500 | @end deffn | |
501 | ||
502 | @deffn {Scheme Procedure} broadcast-condition-variable condvar | |
503 | @deffnx {C Function} scm_broadcast_condition_variable (condvar) | |
504 | Wake up all threads that are waiting for @var{condvar}. | |
505 | @end deffn | |
506 | ||
507 | @sp 1 | |
508 | The following are higher level operations on mutexes. These are | |
509 | available from | |
510 | ||
511 | @example | |
512 | (use-modules (ice-9 threads)) | |
513 | @end example | |
514 | ||
515 | @deffn macro with-mutex mutex [body@dots{}] | |
516 | Lock @var{mutex}, evaluate the @var{body} forms, then unlock | |
517 | @var{mutex}. The return value is the return from the last @var{body} | |
518 | form. | |
519 | ||
520 | The lock, body and unlock form the branches of a @code{dynamic-wind} | |
521 | (@pxref{Dynamic Wind}), so @var{mutex} is automatically unlocked if an | |
522 | error or new continuation exits @var{body}, and is re-locked if | |
523 | @var{body} is re-entered by a captured continuation. | |
524 | @end deffn | |
525 | ||
526 | @deffn macro monitor body@dots{} | |
527 | Evaluate the @var{body} forms, with a mutex locked so only one thread | |
528 | can execute that code at any one time. The return value is the return | |
529 | from the last @var{body} form. | |
530 | ||
531 | Each @code{monitor} form has its own private mutex and the locking and | |
532 | evaluation is as per @code{with-mutex} above. A standard mutex | |
533 | (@code{make-mutex}) is used, which means @var{body} must not | |
534 | recursively re-enter the @code{monitor} form. | |
535 | ||
536 | The term ``monitor'' comes from operating system theory, where it | |
537 | means a particular bit of code managing access to some resource and | |
538 | which only ever executes on behalf of one process at any one time. | |
539 | @end deffn | |
540 | ||
541 | ||
b4fddbbe MV |
542 | @node Blocking |
543 | @subsection Blocking in Guile Mode | |
07d83abe | 544 | |
d3d66147 LC |
545 | Up to Guile version 1.8, a thread blocked in guile mode would prevent |
546 | the garbage collector from running. Thus threads had to explicitly | |
547 | leave guile mode with @code{scm_without_guile ()} before making a | |
548 | potentially blocking call such as a mutex lock, a @code{select ()} | |
549 | system call, etc. The following functions could be used to temporarily | |
550 | leave guile mode or to perform some common blocking operations in a | |
551 | supported way. | |
552 | ||
553 | Starting from Guile 2.0, blocked threads no longer hinder garbage | |
554 | collection. Thus, the functions below are not needed anymore. They can | |
555 | still be used to inform the GC that a thread is about to block, giving | |
556 | it a (small) optimization opportunity for ``stop the world'' garbage | |
557 | collections, should they occur while the thread is blocked. | |
07d83abe | 558 | |
54428bb8 MV |
559 | @deftypefn {C Function} {void *} scm_without_guile (void *(*func) (void *), void *data) |
560 | Leave guile mode, call @var{func} on @var{data}, enter guile mode and | |
561 | return the result of calling @var{func}. | |
07d83abe | 562 | |
b4fddbbe | 563 | While a thread has left guile mode, it must not call any libguile |
54428bb8 MV |
564 | functions except @code{scm_with_guile} or @code{scm_without_guile} and |
565 | must not use any libguile macros. Also, local variables of type | |
566 | @code{SCM} that are allocated while not in guile mode are not | |
567 | protected from the garbage collector. | |
568 | ||
569 | When used from non-guile mode, calling @code{scm_without_guile} is | |
570 | still allowed: it simply calls @var{func}. In that way, you can leave | |
571 | guile mode without having to know whether the current thread is in | |
572 | guile mode or not. | |
07d83abe MV |
573 | @end deftypefn |
574 | ||
b4fddbbe MV |
575 | @deftypefn {C Function} int scm_pthread_mutex_lock (pthread_mutex_t *mutex) |
576 | Like @code{pthread_mutex_lock}, but leaves guile mode while waiting for | |
577 | the mutex. | |
07d83abe MV |
578 | @end deftypefn |
579 | ||
b4fddbbe MV |
580 | @deftypefn {C Function} int scm_pthread_cond_wait (pthread_cond_t *cond, pthread_mutex_t *mutex) |
581 | @deftypefnx {C Function} int scm_pthread_cond_timedwait (pthread_cond_t *cond, pthread_mutex_t *mutex, struct timespec *abstime) | |
582 | Like @code{pthread_cond_wait} and @code{pthread_cond_timedwait}, but | |
583 | leaves guile mode while waiting for the condition variable. | |
07d83abe MV |
584 | @end deftypefn |
585 | ||
b4fddbbe MV |
586 | @deftypefn {C Function} int scm_std_select (int nfds, fd_set *readfds, fd_set *writefds, fd_set *exceptfds, struct timeval *timeout) |
587 | Like @code{select} but leaves guile mode while waiting. Also, the | |
588 | delivery of a system async causes this function to be interrupted with | |
589 | error code @code{EINTR}. | |
07d83abe MV |
590 | @end deftypefn |
591 | ||
b4fddbbe MV |
592 | @deftypefn {C Function} {unsigned int} scm_std_sleep ({unsigned int} seconds) |
593 | Like @code{sleep}, but leaves guile mode while sleeping. Also, the | |
594 | delivery of a system async causes this function to be interrupted. | |
07d83abe MV |
595 | @end deftypefn |
596 | ||
b4fddbbe MV |
597 | @deftypefn {C Function} {unsigned long} scm_std_usleep ({unsigned long} usecs) |
598 | Like @code{usleep}, but leaves guile mode while sleeping. Also, the | |
599 | delivery of a system async causes this function to be interrupted. | |
07d83abe MV |
600 | @end deftypefn |
601 | ||
07d83abe | 602 | |
2567692a MV |
603 | @node Critical Sections |
604 | @subsection Critical Sections | |
605 | ||
606 | @deffn {C Macro} SCM_CRITICAL_SECTION_START | |
607 | @deffnx {C Macro} SCM_CRITICAL_SECTION_END | |
608 | These two macros can be used to delimit a critical section. | |
609 | Syntactically, they are both statements and need to be followed | |
610 | immediately by a semicolon. | |
611 | ||
612 | Executing @code{SCM_CRITICAL_SECTION_START} will lock a recursive | |
613 | mutex and block the executing of system asyncs. Executing | |
614 | @code{SCM_CRITICAL_SECTION_END} will unblock the execution of system | |
615 | asyncs and unlock the mutex. Thus, the code that executes between | |
616 | these two macros can only be executed in one thread at any one time | |
617 | and no system asyncs will run. However, because the mutex is a | |
618 | recursive one, the code might still be reentered by the same thread. | |
619 | You must either allow for this or avoid it, both by careful coding. | |
620 | ||
621 | On the other hand, critical sections delimited with these macros can | |
622 | be nested since the mutex is recursive. | |
623 | ||
624 | You must make sure that for each @code{SCM_CRITICAL_SECTION_START}, | |
625 | the corresponding @code{SCM_CRITICAL_SECTION_END} is always executed. | |
626 | This means that no non-local exit (such as a signalled error) might | |
627 | happen, for example. | |
628 | @end deffn | |
629 | ||
661ae7ab MV |
630 | @deftypefn {C Function} void scm_dynwind_critical_section (SCM mutex) |
631 | Call @code{scm_dynwind_lock_mutex} on @var{mutex} and call | |
632 | @code{scm_dynwind_block_asyncs}. When @var{mutex} is false, a recursive | |
2567692a MV |
633 | mutex provided by Guile is used instead. |
634 | ||
661ae7ab MV |
635 | The effect of a call to @code{scm_dynwind_critical_section} is that |
636 | the current dynwind context (@pxref{Dynamic Wind}) turns into a | |
637 | critical section. Because of the locked mutex, no second thread can | |
638 | enter it concurrently and because of the blocked asyncs, no system | |
639 | async can reenter it from the current thread. | |
2567692a MV |
640 | |
641 | When the current thread reenters the critical section anyway, the kind | |
642 | of @var{mutex} determines what happens: When @var{mutex} is recursive, | |
643 | the reentry is allowed. When it is a normal mutex, an error is | |
644 | signalled. | |
645 | @end deftypefn | |
646 | ||
647 | ||
b4fddbbe MV |
648 | @node Fluids and Dynamic States |
649 | @subsection Fluids and Dynamic States | |
07d83abe MV |
650 | |
651 | @cindex fluids | |
652 | ||
b4fddbbe MV |
653 | A @emph{fluid} is an object that can store one value per @emph{dynamic |
654 | state}. Each thread has a current dynamic state, and when accessing a | |
655 | fluid, this current dynamic state is used to provide the actual value. | |
656 | In this way, fluids can be used for thread local storage, but they are | |
657 | in fact more flexible: dynamic states are objects of their own and can | |
658 | be made current for more than one thread at the same time, or only be | |
659 | made current temporarily, for example. | |
660 | ||
661 | Fluids can also be used to simulate the desirable effects of | |
662 | dynamically scoped variables. Dynamically scoped variables are useful | |
663 | when you want to set a variable to a value during some dynamic extent | |
664 | in the execution of your program and have them revert to their | |
665 | original value when the control flow is outside of this dynamic | |
666 | extent. See the description of @code{with-fluids} below for details. | |
07d83abe MV |
667 | |
668 | New fluids are created with @code{make-fluid} and @code{fluid?} is | |
669 | used for testing whether an object is actually a fluid. The values | |
670 | stored in a fluid can be accessed with @code{fluid-ref} and | |
671 | @code{fluid-set!}. | |
672 | ||
673 | @deffn {Scheme Procedure} make-fluid | |
674 | @deffnx {C Function} scm_make_fluid () | |
675 | Return a newly created fluid. | |
b4fddbbe MV |
676 | Fluids are objects that can hold one |
677 | value per dynamic state. That is, modifications to this value are | |
678 | only visible to code that executes with the same dynamic state as | |
679 | the modifying code. When a new dynamic state is constructed, it | |
680 | inherits the values from its parent. Because each thread normally executes | |
681 | with its own dynamic state, you can use fluids for thread local storage. | |
07d83abe MV |
682 | @end deffn |
683 | ||
684 | @deffn {Scheme Procedure} fluid? obj | |
685 | @deffnx {C Function} scm_fluid_p (obj) | |
686 | Return @code{#t} iff @var{obj} is a fluid; otherwise, return | |
687 | @code{#f}. | |
688 | @end deffn | |
689 | ||
690 | @deffn {Scheme Procedure} fluid-ref fluid | |
691 | @deffnx {C Function} scm_fluid_ref (fluid) | |
692 | Return the value associated with @var{fluid} in the current | |
693 | dynamic root. If @var{fluid} has not been set, then return | |
694 | @code{#f}. | |
695 | @end deffn | |
696 | ||
697 | @deffn {Scheme Procedure} fluid-set! fluid value | |
698 | @deffnx {C Function} scm_fluid_set_x (fluid, value) | |
699 | Set the value associated with @var{fluid} in the current dynamic root. | |
700 | @end deffn | |
701 | ||
702 | @code{with-fluids*} temporarily changes the values of one or more fluids, | |
703 | so that the given procedure and each procedure called by it access the | |
704 | given values. After the procedure returns, the old values are restored. | |
705 | ||
cdf1ad3b MV |
706 | @deffn {Scheme Procedure} with-fluid* fluid value thunk |
707 | @deffnx {C Function} scm_with_fluid (fluid, value, thunk) | |
708 | Set @var{fluid} to @var{value} temporarily, and call @var{thunk}. | |
709 | @var{thunk} must be a procedure with no argument. | |
710 | @end deffn | |
711 | ||
07d83abe MV |
712 | @deffn {Scheme Procedure} with-fluids* fluids values thunk |
713 | @deffnx {C Function} scm_with_fluids (fluids, values, thunk) | |
714 | Set @var{fluids} to @var{values} temporary, and call @var{thunk}. | |
715 | @var{fluids} must be a list of fluids and @var{values} must be the | |
716 | same number of their values to be applied. Each substitution is done | |
717 | in the order given. @var{thunk} must be a procedure with no argument. | |
718 | it is called inside a @code{dynamic-wind} and the fluids are | |
719 | set/restored when control enter or leaves the established dynamic | |
720 | extent. | |
721 | @end deffn | |
722 | ||
723 | @deffn {Scheme Macro} with-fluids ((fluid value) ...) body... | |
724 | Execute @var{body...} while each @var{fluid} is set to the | |
725 | corresponding @var{value}. Both @var{fluid} and @var{value} are | |
726 | evaluated and @var{fluid} must yield a fluid. @var{body...} is | |
727 | executed inside a @code{dynamic-wind} and the fluids are set/restored | |
728 | when control enter or leaves the established dynamic extent. | |
729 | @end deffn | |
730 | ||
731 | @deftypefn {C Function} SCM scm_c_with_fluids (SCM fluids, SCM vals, SCM (*cproc)(void *), void *data) | |
732 | @deftypefnx {C Function} SCM scm_c_with_fluid (SCM fluid, SCM val, SCM (*cproc)(void *), void *data) | |
733 | The function @code{scm_c_with_fluids} is like @code{scm_with_fluids} | |
734 | except that it takes a C function to call instead of a Scheme thunk. | |
735 | ||
736 | The function @code{scm_c_with_fluid} is similar but only allows one | |
737 | fluid to be set instead of a list. | |
738 | @end deftypefn | |
739 | ||
661ae7ab | 740 | @deftypefn {C Function} void scm_dynwind_fluid (SCM fluid, SCM val) |
07d83abe | 741 | This function must be used inside a pair of calls to |
661ae7ab MV |
742 | @code{scm_dynwind_begin} and @code{scm_dynwind_end} (@pxref{Dynamic |
743 | Wind}). During the dynwind context, the fluid @var{fluid} is set to | |
744 | @var{val}. | |
07d83abe MV |
745 | |
746 | More precisely, the value of the fluid is swapped with a `backup' | |
661ae7ab MV |
747 | value whenever the dynwind context is entered or left. The backup |
748 | value is initialized with the @var{val} argument. | |
07d83abe MV |
749 | @end deftypefn |
750 | ||
b4fddbbe MV |
751 | @deffn {Scheme Procedure} make-dynamic-state [parent] |
752 | @deffnx {C Function} scm_make_dynamic_state (parent) | |
753 | Return a copy of the dynamic state object @var{parent} | |
754 | or of the current dynamic state when @var{parent} is omitted. | |
755 | @end deffn | |
756 | ||
757 | @deffn {Scheme Procedure} dynamic-state? obj | |
758 | @deffnx {C Function} scm_dynamic_state_p (obj) | |
759 | Return @code{#t} if @var{obj} is a dynamic state object; | |
760 | return @code{#f} otherwise. | |
761 | @end deffn | |
762 | ||
763 | @deftypefn {C Procedure} int scm_is_dynamic_state (SCM obj) | |
764 | Return non-zero if @var{obj} is a dynamic state object; | |
765 | return zero otherwise. | |
766 | @end deftypefn | |
767 | ||
768 | @deffn {Scheme Procedure} current-dynamic-state | |
769 | @deffnx {C Function} scm_current_dynamic_state () | |
770 | Return the current dynamic state object. | |
771 | @end deffn | |
772 | ||
773 | @deffn {Scheme Procedure} set-current-dynamic-state state | |
774 | @deffnx {C Function} scm_set_current_dynamic_state (state) | |
775 | Set the current dynamic state object to @var{state} | |
776 | and return the previous current dynamic state object. | |
777 | @end deffn | |
778 | ||
779 | @deffn {Scheme Procedure} with-dynamic-state state proc | |
780 | @deffnx {C Function} scm_with_dynamic_state (state, proc) | |
781 | Call @var{proc} while @var{state} is the current dynamic | |
782 | state object. | |
783 | @end deffn | |
784 | ||
661ae7ab MV |
785 | @deftypefn {C Procedure} void scm_dynwind_current_dynamic_state (SCM state) |
786 | Set the current dynamic state to @var{state} for the current dynwind | |
787 | context. | |
b4fddbbe MV |
788 | @end deftypefn |
789 | ||
c2110081 | 790 | @deftypefn {C Procedure} {void *} scm_c_with_dynamic_state (SCM state, void *(*func)(void *), void *data) |
b4fddbbe MV |
791 | Like @code{scm_with_dynamic_state}, but call @var{func} with |
792 | @var{data}. | |
793 | @end deftypefn | |
794 | ||
07d83abe MV |
795 | @node Parallel Forms |
796 | @subsection Parallel forms | |
797 | @cindex parallel forms | |
798 | ||
799 | The functions described in this section are available from | |
800 | ||
801 | @example | |
802 | (use-modules (ice-9 threads)) | |
803 | @end example | |
804 | ||
805 | @deffn syntax parallel expr1 @dots{} exprN | |
af1323c5 | 806 | Evaluate each @var{expr} expression in parallel, each in its own thread. |
07d83abe MV |
807 | Return the results as a set of @var{N} multiple values |
808 | (@pxref{Multiple Values}). | |
809 | @end deffn | |
810 | ||
811 | @deffn syntax letpar ((var1 expr1) @dots{} (varN exprN)) body@dots{} | |
af1323c5 | 812 | Evaluate each @var{expr} in parallel, each in its own thread, then bind |
07d83abe MV |
813 | the results to the corresponding @var{var} variables and evaluate |
814 | @var{body}. | |
815 | ||
816 | @code{letpar} is like @code{let} (@pxref{Local Bindings}), but all the | |
817 | expressions for the bindings are evaluated in parallel. | |
818 | @end deffn | |
819 | ||
820 | @deffn {Scheme Procedure} par-map proc lst1 @dots{} lstN | |
821 | @deffnx {Scheme Procedure} par-for-each proc lst1 @dots{} lstN | |
822 | Call @var{proc} on the elements of the given lists. @code{par-map} | |
823 | returns a list comprising the return values from @var{proc}. | |
824 | @code{par-for-each} returns an unspecified value, but waits for all | |
825 | calls to complete. | |
826 | ||
827 | The @var{proc} calls are @code{(@var{proc} @var{elem1} @dots{} | |
828 | @var{elemN})}, where each @var{elem} is from the corresponding | |
829 | @var{lst}. Each @var{lst} must be the same length. The calls are | |
af1323c5 | 830 | made in parallel, each in its own thread. |
07d83abe MV |
831 | |
832 | These functions are like @code{map} and @code{for-each} (@pxref{List | |
833 | Mapping}), but make their @var{proc} calls in parallel. | |
834 | @end deffn | |
835 | ||
836 | @deffn {Scheme Procedure} n-par-map n proc lst1 @dots{} lstN | |
837 | @deffnx {Scheme Procedure} n-par-for-each n proc lst1 @dots{} lstN | |
838 | Call @var{proc} on the elements of the given lists, in the same way as | |
839 | @code{par-map} and @code{par-for-each} above, but use no more than | |
af1323c5 | 840 | @var{n} threads at any one time. The order in which calls are |
07d83abe MV |
841 | initiated within that threads limit is unspecified. |
842 | ||
843 | These functions are good for controlling resource consumption if | |
844 | @var{proc} calls might be costly, or if there are many to be made. On | |
845 | a dual-CPU system for instance @math{@var{n}=4} might be enough to | |
846 | keep the CPUs utilized, and not consume too much memory. | |
847 | @end deffn | |
848 | ||
849 | @deffn {Scheme Procedure} n-for-each-par-map n sproc pproc lst1 @dots{} lstN | |
850 | Apply @var{pproc} to the elements of the given lists, and apply | |
851 | @var{sproc} to each result returned by @var{pproc}. The final return | |
852 | value is unspecified, but all calls will have been completed before | |
853 | returning. | |
854 | ||
855 | The calls made are @code{(@var{sproc} (@var{pproc} @var{elem1} @dots{} | |
856 | @var{elemN}))}, where each @var{elem} is from the corresponding | |
857 | @var{lst}. Each @var{lst} must have the same number of elements. | |
858 | ||
af1323c5 KR |
859 | The @var{pproc} calls are made in parallel, in separate threads. No more |
860 | than @var{n} threads are used at any one time. The order in which | |
07d83abe MV |
861 | @var{pproc} calls are initiated within that limit is unspecified. |
862 | ||
863 | The @var{sproc} calls are made serially, in list element order, one at | |
864 | a time. @var{pproc} calls on later elements may execute in parallel | |
865 | with the @var{sproc} calls. Exactly which thread makes each | |
866 | @var{sproc} call is unspecified. | |
867 | ||
868 | This function is designed for individual calculations that can be done | |
869 | in parallel, but with results needing to be handled serially, for | |
870 | instance to write them to a file. The @var{n} limit on threads | |
871 | controls system resource usage when there are many calculations or | |
872 | when they might be costly. | |
873 | ||
874 | It will be seen that @code{n-for-each-par-map} is like a combination | |
875 | of @code{n-par-map} and @code{for-each}, | |
876 | ||
877 | @example | |
af1323c5 | 878 | (for-each sproc (n-par-map n pproc lst1 ... lstN)) |
07d83abe MV |
879 | @end example |
880 | ||
881 | @noindent | |
882 | But the actual implementation is more efficient since each @var{sproc} | |
883 | call, in turn, can be initiated once the relevant @var{pproc} call has | |
884 | completed, it doesn't need to wait for all to finish. | |
885 | @end deffn | |
886 | ||
887 | ||
3cf066df | 888 | |
07d83abe MV |
889 | @c Local Variables: |
890 | @c TeX-master: "guile.texi" | |
891 | @c End: |