1 /* ---------------------------------------------------------------------------
2 * $Id: Schedule.c,v 1.97 2001/07/23 17:23:19 simonmar Exp $
4 * (c) The GHC Team, 1998-2000
8 * Different GHC ways use this scheduler quite differently (see comments below)
9 * Here is the global picture:
11 * WAY Name CPP flag What's it for
12 * --------------------------------------
13 * mp GUM PAR Parallel execution on a distributed memory machine
14 * s SMP SMP Parallel execution on a shared memory machine
15 * mg GranSim GRAN Simulation of parallel execution
16 * md GUM/GdH DIST Distributed execution (based on GUM)
17 * --------------------------------------------------------------------------*/
19 //@node Main scheduling code, , ,
20 //@section Main scheduling code
23 * Version with scheduler monitor support for SMPs (WAY=s):
25 This design provides a high-level API to create and schedule threads etc.
26 as documented in the SMP design document.
28 It uses a monitor design controlled by a single mutex to exercise control
29 over accesses to shared data structures, and builds on the Posix threads
32 The majority of state is shared. In order to keep essential per-task state,
33 there is a Capability structure, which contains all the information
34 needed to run a thread: its STG registers, a pointer to its TSO, a
35 nursery etc. During STG execution, a pointer to the capability is
36 kept in a register (BaseReg).
38 In a non-SMP build, there is one global capability, namely MainRegTable.
42 * Version with support for distributed memory parallelism aka GUM (WAY=mp):
44 The main scheduling loop in GUM iterates until a finish message is received.
45 In that case a global flag @receivedFinish@ is set and this instance of
46 the RTS shuts down. See ghc/rts/parallel/HLComms.c:processMessages()
47 for the handling of incoming messages, such as PP_FINISH.
48 Note that in the parallel case we have a system manager that coordinates
49 different PEs, each of which are running one instance of the RTS.
50 See ghc/rts/parallel/SysMan.c for the main routine of the parallel program.
51 From this routine processes executing ghc/rts/Main.c are spawned. -- HWL
53 * Version with support for simulating parallel execution aka GranSim (WAY=mg):
55 The main scheduling code in GranSim is quite different from that in std
56 (concurrent) Haskell: while concurrent Haskell just iterates over the
57 threads in the runnable queue, GranSim is event driven, i.e. it iterates
58 over the events in the global event queue. -- HWL
63 //* Variables and Data structures::
64 //* Main scheduling loop::
65 //* Suspend and Resume::
67 //* Garbage Collextion Routines::
68 //* Blocking Queue Routines::
69 //* Exception Handling Routines::
70 //* Debugging Routines::
74 //@node Includes, Variables and Data structures, Main scheduling code, Main scheduling code
75 //@subsection Includes
83 #include "StgStartup.h"
86 #include "StgMiscClosures.h"
88 #include "Interpreter.h"
89 #include "Exception.h"
97 #if defined(GRAN) || defined(PAR)
98 # include "GranSimRts.h"
100 # include "ParallelRts.h"
101 # include "Parallel.h"
102 # include "ParallelDebug.h"
103 # include "FetchMe.h"
110 //@node Variables and Data structures, Prototypes, Includes, Main scheduling code
111 //@subsection Variables and Data structures
115 * These are the threads which clients have requested that we run.
117 * In an SMP build, we might have several concurrent clients all
118 * waiting for results, and each one will wait on a condition variable
119 * until the result is available.
121 * In non-SMP, clients are strictly nested: the first client calls
122 * into the RTS, which might call out again to C with a _ccall_GC, and
123 * eventually re-enter the RTS.
125 * Main threads information is kept in a linked list:
127 //@cindex StgMainThread
128 typedef struct StgMainThread_ {
130 SchedulerStatus stat;
133 pthread_cond_t wakeup;
135 struct StgMainThread_ *link;
138 /* Main thread queue.
139 * Locks required: sched_mutex.
141 static StgMainThread *main_threads;
144 * Locks required: sched_mutex.
148 StgTSO* ActiveTSO = NULL; /* for assigning system costs; GranSim-Light only */
149 /* rtsTime TimeOfNextEvent, EndOfTimeSlice; now in GranSim.c */
152 In GranSim we have a runable and a blocked queue for each processor.
153 In order to minimise code changes new arrays run_queue_hds/tls
154 are created. run_queue_hd is then a short cut (macro) for
155 run_queue_hds[CurrentProc] (see GranSim.h).
158 StgTSO *run_queue_hds[MAX_PROC], *run_queue_tls[MAX_PROC];
159 StgTSO *blocked_queue_hds[MAX_PROC], *blocked_queue_tls[MAX_PROC];
160 StgTSO *ccalling_threadss[MAX_PROC];
161 /* We use the same global list of threads (all_threads) in GranSim as in
162 the std RTS (i.e. we are cheating). However, we don't use this list in
163 the GranSim specific code at the moment (so we are only potentially
168 StgTSO *run_queue_hd, *run_queue_tl;
169 StgTSO *blocked_queue_hd, *blocked_queue_tl;
170 StgTSO *sleeping_queue; /* perhaps replace with a hash table? */
174 /* Linked list of all threads.
175 * Used for detecting garbage collected threads.
179 /* Threads suspended in _ccall_GC.
181 static StgTSO *suspended_ccalling_threads;
183 static void GetRoots(evac_fn);
184 static StgTSO *threadStackOverflow(StgTSO *tso);
186 /* KH: The following two flags are shared memory locations. There is no need
187 to lock them, since they are only unset at the end of a scheduler
191 /* flag set by signal handler to precipitate a context switch */
192 //@cindex context_switch
195 /* if this flag is set as well, give up execution */
196 //@cindex interrupted
199 /* Next thread ID to allocate.
200 * Locks required: sched_mutex
202 //@cindex next_thread_id
203 StgThreadID next_thread_id = 1;
206 * Pointers to the state of the current thread.
207 * Rule of thumb: if CurrentTSO != NULL, then we're running a Haskell
208 * thread. If CurrentTSO == NULL, then we're at the scheduler level.
211 /* The smallest stack size that makes any sense is:
212 * RESERVED_STACK_WORDS (so we can get back from the stack overflow)
213 * + sizeofW(StgStopFrame) (the stg_stop_thread_info frame)
214 * + 1 (the realworld token for an IO thread)
215 * + 1 (the closure to enter)
217 * A thread with this stack will bomb immediately with a stack
218 * overflow, which will increase its stack size.
221 #define MIN_STACK_WORDS (RESERVED_STACK_WORDS + sizeofW(StgStopFrame) + 2)
223 /* Free capability list.
224 * Locks required: sched_mutex.
227 //@cindex free_capabilities
228 //@cindex n_free_capabilities
229 Capability *free_capabilities; /* Available capabilities for running threads */
230 nat n_free_capabilities; /* total number of available capabilities */
232 //@cindex MainRegTable
233 Capability MainRegTable; /* for non-SMP, we have one global capability */
240 /* This is used in `TSO.h' and gcc 2.96 insists that this variable actually
241 * exists - earlier gccs apparently didn't.
248 /* All our current task ids, saved in case we need to kill them later.
255 void addToBlockedQueue ( StgTSO *tso );
257 static void schedule ( void );
258 void interruptStgRts ( void );
260 static StgTSO * createThread_ ( nat size, rtsBool have_lock, StgInt pri );
262 static StgTSO * createThread_ ( nat size, rtsBool have_lock );
265 static void detectBlackHoles ( void );
268 static void sched_belch(char *s, ...);
272 //@cindex sched_mutex
274 //@cindex thread_ready_cond
275 //@cindex gc_pending_cond
276 pthread_mutex_t sched_mutex = PTHREAD_MUTEX_INITIALIZER;
277 pthread_mutex_t term_mutex = PTHREAD_MUTEX_INITIALIZER;
278 pthread_cond_t thread_ready_cond = PTHREAD_COND_INITIALIZER;
279 pthread_cond_t gc_pending_cond = PTHREAD_COND_INITIALIZER;
286 rtsTime TimeOfLastYield;
287 rtsBool emitSchedule = rtsTrue;
291 char *whatNext_strs[] = {
299 char *threadReturnCode_strs[] = {
300 "HeapOverflow", /* might also be StackOverflow */
309 StgTSO * createSparkThread(rtsSpark spark);
310 StgTSO * activateSpark (rtsSpark spark);
314 * The thread state for the main thread.
315 // ToDo: check whether not needed any more
319 //@node Main scheduling loop, Suspend and Resume, Prototypes, Main scheduling code
320 //@subsection Main scheduling loop
322 /* ---------------------------------------------------------------------------
323 Main scheduling loop.
325 We use round-robin scheduling, each thread returning to the
326 scheduler loop when one of these conditions is detected:
329 * timer expires (thread yields)
334 Locking notes: we acquire the scheduler lock once at the beginning
335 of the scheduler loop, and release it when
337 * running a thread, or
338 * waiting for work, or
339 * waiting for a GC to complete.
342 In a GranSim setup this loop iterates over the global event queue.
343 This revolves around the global event queue, which determines what
344 to do next. Therefore, it's more complicated than either the
345 concurrent or the parallel (GUM) setup.
348 GUM iterates over incoming messages.
349 It starts with nothing to do (thus CurrentTSO == END_TSO_QUEUE),
350 and sends out a fish whenever it has nothing to do; in-between
351 doing the actual reductions (shared code below) it processes the
352 incoming messages and deals with delayed operations
353 (see PendingFetches).
354 This is not the ugliest code you could imagine, but it's bloody close.
356 ------------------------------------------------------------------------ */
363 StgThreadReturnCode ret;
371 rtsBool receivedFinish = rtsFalse;
373 nat tp_size, sp_size; // stats only
376 rtsBool was_interrupted = rtsFalse;
378 ACQUIRE_LOCK(&sched_mutex);
382 /* set up first event to get things going */
383 /* ToDo: assign costs for system setup and init MainTSO ! */
384 new_event(CurrentProc, CurrentProc, CurrentTime[CurrentProc],
386 CurrentTSO, (StgClosure*)NULL, (rtsSpark*)NULL);
389 fprintf(stderr, "GRAN: Init CurrentTSO (in schedule) = %p\n", CurrentTSO);
390 G_TSO(CurrentTSO, 5));
392 if (RtsFlags.GranFlags.Light) {
393 /* Save current time; GranSim Light only */
394 CurrentTSO->gran.clock = CurrentTime[CurrentProc];
397 event = get_next_event();
399 while (event!=(rtsEvent*)NULL) {
400 /* Choose the processor with the next event */
401 CurrentProc = event->proc;
402 CurrentTSO = event->tso;
406 while (!receivedFinish) { /* set by processMessages */
407 /* when receiving PP_FINISH message */
414 IF_DEBUG(scheduler, printAllThreads());
416 /* If we're interrupted (the user pressed ^C, or some other
417 * termination condition occurred), kill all the currently running
421 IF_DEBUG(scheduler, sched_belch("interrupted"));
423 interrupted = rtsFalse;
424 was_interrupted = rtsTrue;
427 /* Go through the list of main threads and wake up any
428 * clients whose computations have finished. ToDo: this
429 * should be done more efficiently without a linear scan
430 * of the main threads list, somehow...
434 StgMainThread *m, **prev;
435 prev = &main_threads;
436 for (m = main_threads; m != NULL; m = m->link) {
437 switch (m->tso->what_next) {
440 *(m->ret) = (StgClosure *)m->tso->sp[0];
444 pthread_cond_broadcast(&m->wakeup);
448 if (was_interrupted) {
449 m->stat = Interrupted;
453 pthread_cond_broadcast(&m->wakeup);
463 /* in GUM do this only on the Main PE */
466 /* If our main thread has finished or been killed, return.
469 StgMainThread *m = main_threads;
470 if (m->tso->what_next == ThreadComplete
471 || m->tso->what_next == ThreadKilled) {
472 main_threads = main_threads->link;
473 if (m->tso->what_next == ThreadComplete) {
474 /* we finished successfully, fill in the return value */
475 if (m->ret) { *(m->ret) = (StgClosure *)m->tso->sp[0]; };
479 if (was_interrupted) {
480 m->stat = Interrupted;
490 /* Top up the run queue from our spark pool. We try to make the
491 * number of threads in the run queue equal to the number of
496 nat n = n_free_capabilities;
497 StgTSO *tso = run_queue_hd;
499 /* Count the run queue */
500 while (n > 0 && tso != END_TSO_QUEUE) {
507 spark = findSpark(rtsFalse);
509 break; /* no more sparks in the pool */
511 /* I'd prefer this to be done in activateSpark -- HWL */
512 /* tricky - it needs to hold the scheduler lock and
513 * not try to re-acquire it -- SDM */
514 createSparkThread(spark);
516 sched_belch("==^^ turning spark of closure %p into a thread",
517 (StgClosure *)spark));
520 /* We need to wake up the other tasks if we just created some
523 if (n_free_capabilities - n > 1) {
524 pthread_cond_signal(&thread_ready_cond);
529 /* Check whether any waiting threads need to be woken up. If the
530 * run queue is empty, and there are no other tasks running, we
531 * can wait indefinitely for something to happen.
532 * ToDo: what if another client comes along & requests another
535 if (blocked_queue_hd != END_TSO_QUEUE || sleeping_queue != END_TSO_QUEUE) {
537 (run_queue_hd == END_TSO_QUEUE)
539 && (n_free_capabilities == RtsFlags.ParFlags.nNodes)
543 /* we can be interrupted while waiting for I/O... */
544 if (interrupted) continue;
546 /* check for signals each time around the scheduler */
547 #ifndef mingw32_TARGET_OS
548 if (signals_pending()) {
549 start_signal_handlers();
554 * Detect deadlock: when we have no threads to run, there are no
555 * threads waiting on I/O or sleeping, and all the other tasks are
556 * waiting for work, we must have a deadlock of some description.
558 * We first try to find threads blocked on themselves (ie. black
559 * holes), and generate NonTermination exceptions where necessary.
561 * If no threads are black holed, we have a deadlock situation, so
562 * inform all the main threads.
565 if (blocked_queue_hd == END_TSO_QUEUE
566 && run_queue_hd == END_TSO_QUEUE
567 && sleeping_queue == END_TSO_QUEUE
569 && (n_free_capabilities == RtsFlags.ParFlags.nNodes)
573 IF_DEBUG(scheduler, sched_belch("deadlocked, forcing major GC..."));
574 GarbageCollect(GetRoots,rtsTrue);
575 if (blocked_queue_hd == END_TSO_QUEUE
576 && run_queue_hd == END_TSO_QUEUE
577 && sleeping_queue == END_TSO_QUEUE) {
578 IF_DEBUG(scheduler, sched_belch("still deadlocked, checking for black holes..."));
580 if (run_queue_hd == END_TSO_QUEUE) {
581 StgMainThread *m = main_threads;
583 for (; m != NULL; m = m->link) {
586 pthread_cond_broadcast(&m->wakeup);
592 main_threads = m->link;
599 /* ToDo: add deadlock detection in GUM (similar to SMP) -- HWL */
603 /* If there's a GC pending, don't do anything until it has
607 IF_DEBUG(scheduler,sched_belch("waiting for GC"));
608 pthread_cond_wait(&gc_pending_cond, &sched_mutex);
611 /* block until we've got a thread on the run queue and a free
614 while (run_queue_hd == END_TSO_QUEUE || free_capabilities == NULL) {
615 IF_DEBUG(scheduler, sched_belch("waiting for work"));
616 pthread_cond_wait(&thread_ready_cond, &sched_mutex);
617 IF_DEBUG(scheduler, sched_belch("work now available"));
623 if (RtsFlags.GranFlags.Light)
624 GranSimLight_enter_system(event, &ActiveTSO); // adjust ActiveTSO etc
626 /* adjust time based on time-stamp */
627 if (event->time > CurrentTime[CurrentProc] &&
628 event->evttype != ContinueThread)
629 CurrentTime[CurrentProc] = event->time;
631 /* Deal with the idle PEs (may issue FindWork or MoveSpark events) */
632 if (!RtsFlags.GranFlags.Light)
635 IF_DEBUG(gran, fprintf(stderr, "GRAN: switch by event-type\n"));
637 /* main event dispatcher in GranSim */
638 switch (event->evttype) {
639 /* Should just be continuing execution */
641 IF_DEBUG(gran, fprintf(stderr, "GRAN: doing ContinueThread\n"));
642 /* ToDo: check assertion
643 ASSERT(run_queue_hd != (StgTSO*)NULL &&
644 run_queue_hd != END_TSO_QUEUE);
646 /* Ignore ContinueThreads for fetching threads (if synchr comm) */
647 if (!RtsFlags.GranFlags.DoAsyncFetch &&
648 procStatus[CurrentProc]==Fetching) {
649 belch("ghuH: Spurious ContinueThread while Fetching ignored; TSO %d (%p) [PE %d]",
650 CurrentTSO->id, CurrentTSO, CurrentProc);
653 /* Ignore ContinueThreads for completed threads */
654 if (CurrentTSO->what_next == ThreadComplete) {
655 belch("ghuH: found a ContinueThread event for completed thread %d (%p) [PE %d] (ignoring ContinueThread)",
656 CurrentTSO->id, CurrentTSO, CurrentProc);
659 /* Ignore ContinueThreads for threads that are being migrated */
660 if (PROCS(CurrentTSO)==Nowhere) {
661 belch("ghuH: trying to run the migrating TSO %d (%p) [PE %d] (ignoring ContinueThread)",
662 CurrentTSO->id, CurrentTSO, CurrentProc);
665 /* The thread should be at the beginning of the run queue */
666 if (CurrentTSO!=run_queue_hds[CurrentProc]) {
667 belch("ghuH: TSO %d (%p) [PE %d] is not at the start of the run_queue when doing a ContinueThread",
668 CurrentTSO->id, CurrentTSO, CurrentProc);
669 break; // run the thread anyway
672 new_event(proc, proc, CurrentTime[proc],
674 (StgTSO*)NULL, (StgClosure*)NULL, (rtsSpark*)NULL);
676 */ /* Catches superfluous CONTINUEs -- should be unnecessary */
677 break; // now actually run the thread; DaH Qu'vam yImuHbej
680 do_the_fetchnode(event);
681 goto next_thread; /* handle next event in event queue */
684 do_the_globalblock(event);
685 goto next_thread; /* handle next event in event queue */
688 do_the_fetchreply(event);
689 goto next_thread; /* handle next event in event queue */
691 case UnblockThread: /* Move from the blocked queue to the tail of */
692 do_the_unblock(event);
693 goto next_thread; /* handle next event in event queue */
695 case ResumeThread: /* Move from the blocked queue to the tail of */
696 /* the runnable queue ( i.e. Qu' SImqa'lu') */
697 event->tso->gran.blocktime +=
698 CurrentTime[CurrentProc] - event->tso->gran.blockedat;
699 do_the_startthread(event);
700 goto next_thread; /* handle next event in event queue */
703 do_the_startthread(event);
704 goto next_thread; /* handle next event in event queue */
707 do_the_movethread(event);
708 goto next_thread; /* handle next event in event queue */
711 do_the_movespark(event);
712 goto next_thread; /* handle next event in event queue */
715 do_the_findwork(event);
716 goto next_thread; /* handle next event in event queue */
719 barf("Illegal event type %u\n", event->evttype);
722 /* This point was scheduler_loop in the old RTS */
724 IF_DEBUG(gran, belch("GRAN: after main switch"));
726 TimeOfLastEvent = CurrentTime[CurrentProc];
727 TimeOfNextEvent = get_time_of_next_event();
728 IgnoreEvents=(TimeOfNextEvent==0); // HWL HACK
729 // CurrentTSO = ThreadQueueHd;
731 IF_DEBUG(gran, belch("GRAN: time of next event is: %ld",
734 if (RtsFlags.GranFlags.Light)
735 GranSimLight_leave_system(event, &ActiveTSO);
737 EndOfTimeSlice = CurrentTime[CurrentProc]+RtsFlags.GranFlags.time_slice;
740 belch("GRAN: end of time-slice is %#lx", EndOfTimeSlice));
742 /* in a GranSim setup the TSO stays on the run queue */
744 /* Take a thread from the run queue. */
745 t = POP_RUN_QUEUE(); // take_off_run_queue(t);
748 fprintf(stderr, "GRAN: About to run current thread, which is\n");
751 context_switch = 0; // turned on via GranYield, checking events and time slice
754 DumpGranEvent(GR_SCHEDULE, t));
756 procStatus[CurrentProc] = Busy;
759 if (PendingFetches != END_BF_QUEUE) {
763 /* ToDo: phps merge with spark activation above */
764 /* check whether we have local work and send requests if we have none */
765 if (EMPTY_RUN_QUEUE()) { /* no runnable threads */
766 /* :-[ no local threads => look out for local sparks */
767 /* the spark pool for the current PE */
768 pool = &(MainRegTable.rSparks); // generalise to cap = &MainRegTable
769 if (advisory_thread_count < RtsFlags.ParFlags.maxThreads &&
770 pool->hd < pool->tl) {
772 * ToDo: add GC code check that we really have enough heap afterwards!!
774 * If we're here (no runnable threads) and we have pending
775 * sparks, we must have a space problem. Get enough space
776 * to turn one of those pending sparks into a
780 spark = findSpark(rtsFalse); /* get a spark */
781 if (spark != (rtsSpark) NULL) {
782 tso = activateSpark(spark); /* turn the spark into a thread */
783 IF_PAR_DEBUG(schedule,
784 belch("==== schedule: Created TSO %d (%p); %d threads active",
785 tso->id, tso, advisory_thread_count));
787 if (tso==END_TSO_QUEUE) { /* failed to activate spark->back to loop */
788 belch("==^^ failed to activate spark");
790 } /* otherwise fall through & pick-up new tso */
792 IF_PAR_DEBUG(verbose,
793 belch("==^^ no local sparks (spark pool contains only NFs: %d)",
794 spark_queue_len(pool)));
799 /* If we still have no work we need to send a FISH to get a spark
802 if (EMPTY_RUN_QUEUE()) {
803 /* =8-[ no local sparks => look for work on other PEs */
805 * We really have absolutely no work. Send out a fish
806 * (there may be some out there already), and wait for
807 * something to arrive. We clearly can't run any threads
808 * until a SCHEDULE or RESUME arrives, and so that's what
809 * we're hoping to see. (Of course, we still have to
810 * respond to other types of messages.)
812 TIME now = msTime() /*CURRENT_TIME*/;
813 IF_PAR_DEBUG(verbose,
814 belch("-- now=%ld", now));
815 IF_PAR_DEBUG(verbose,
816 if (outstandingFishes < RtsFlags.ParFlags.maxFishes &&
817 (last_fish_arrived_at!=0 &&
818 last_fish_arrived_at+RtsFlags.ParFlags.fishDelay > now)) {
819 belch("--$$ delaying FISH until %ld (last fish %ld, delay %ld, now %ld)",
820 last_fish_arrived_at+RtsFlags.ParFlags.fishDelay,
821 last_fish_arrived_at,
822 RtsFlags.ParFlags.fishDelay, now);
825 if (outstandingFishes < RtsFlags.ParFlags.maxFishes &&
826 (last_fish_arrived_at==0 ||
827 (last_fish_arrived_at+RtsFlags.ParFlags.fishDelay <= now))) {
828 /* outstandingFishes is set in sendFish, processFish;
829 avoid flooding system with fishes via delay */
831 sendFish(pe, mytid, NEW_FISH_AGE, NEW_FISH_HISTORY,
834 // Global statistics: count no. of fishes
835 if (RtsFlags.ParFlags.ParStats.Global &&
836 RtsFlags.GcFlags.giveStats > NO_GC_STATS) {
837 globalParStats.tot_fish_mess++;
841 receivedFinish = processMessages();
844 } else if (PacketsWaiting()) { /* Look for incoming messages */
845 receivedFinish = processMessages();
848 /* Now we are sure that we have some work available */
849 ASSERT(run_queue_hd != END_TSO_QUEUE);
851 /* Take a thread from the run queue, if we have work */
852 t = POP_RUN_QUEUE(); // take_off_run_queue(END_TSO_QUEUE);
853 IF_DEBUG(sanity,checkTSO(t));
855 /* ToDo: write something to the log-file
856 if (RTSflags.ParFlags.granSimStats && !sameThread)
857 DumpGranEvent(GR_SCHEDULE, RunnableThreadsHd);
861 /* the spark pool for the current PE */
862 pool = &(MainRegTable.rSparks); // generalise to cap = &MainRegTable
865 belch("--=^ %d threads, %d sparks on [%#x]",
866 run_queue_len(), spark_queue_len(pool), CURRENT_PROC));
869 if (0 && RtsFlags.ParFlags.ParStats.Full &&
870 t && LastTSO && t->id != LastTSO->id &&
871 LastTSO->why_blocked == NotBlocked &&
872 LastTSO->what_next != ThreadComplete) {
873 // if previously scheduled TSO not blocked we have to record the context switch
874 DumpVeryRawGranEvent(TimeOfLastYield, CURRENT_PROC, CURRENT_PROC,
875 GR_DESCHEDULE, LastTSO, (StgClosure *)NULL, 0, 0);
878 if (RtsFlags.ParFlags.ParStats.Full &&
879 (emitSchedule /* forced emit */ ||
880 (t && LastTSO && t->id != LastTSO->id))) {
882 we are running a different TSO, so write a schedule event to log file
883 NB: If we use fair scheduling we also have to write a deschedule
884 event for LastTSO; with unfair scheduling we know that the
885 previous tso has blocked whenever we switch to another tso, so
886 we don't need it in GUM for now
888 DumpRawGranEvent(CURRENT_PROC, CURRENT_PROC,
889 GR_SCHEDULE, t, (StgClosure *)NULL, 0, 0);
890 emitSchedule = rtsFalse;
894 #else /* !GRAN && !PAR */
896 /* grab a thread from the run queue
898 ASSERT(run_queue_hd != END_TSO_QUEUE);
900 IF_DEBUG(sanity,checkTSO(t));
907 cap = free_capabilities;
908 free_capabilities = cap->link;
909 n_free_capabilities--;
914 cap->rCurrentTSO = t;
916 /* context switches are now initiated by the timer signal, unless
917 * the user specified "context switch as often as possible", with
920 if (RtsFlags.ConcFlags.ctxtSwitchTicks == 0
921 && (run_queue_hd != END_TSO_QUEUE
922 || blocked_queue_hd != END_TSO_QUEUE
923 || sleeping_queue != END_TSO_QUEUE))
928 RELEASE_LOCK(&sched_mutex);
930 IF_DEBUG(scheduler, sched_belch("-->> Running TSO %ld (%p) %s ...",
931 t->id, t, whatNext_strs[t->what_next]));
933 /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */
934 /* Run the current thread
936 switch (cap->rCurrentTSO->what_next) {
939 /* Thread already finished, return to scheduler. */
940 ret = ThreadFinished;
943 ret = StgRun((StgFunPtr) stg_enterStackTop, cap);
946 ret = StgRun((StgFunPtr) stg_returnToStackTop, cap);
948 case ThreadEnterInterp:
949 ret = interpretBCO(cap);
952 barf("schedule: invalid what_next field");
954 /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */
956 /* Costs for the scheduler are assigned to CCS_SYSTEM */
961 ACQUIRE_LOCK(&sched_mutex);
964 IF_DEBUG(scheduler,fprintf(stderr,"scheduler (task %ld): ", pthread_self()););
965 #elif !defined(GRAN) && !defined(PAR)
966 IF_DEBUG(scheduler,fprintf(stderr,"scheduler: "););
968 t = cap->rCurrentTSO;
971 /* HACK 675: if the last thread didn't yield, make sure to print a
972 SCHEDULE event to the log file when StgRunning the next thread, even
973 if it is the same one as before */
975 TimeOfLastYield = CURRENT_TIME;
982 DumpGranEvent(GR_DESCHEDULE, t));
983 globalGranStats.tot_heapover++;
986 //DumpGranEvent(GR_DESCHEDULE, t);
987 globalParStats.tot_heapover++;
989 /* make all the running tasks block on a condition variable,
990 * maybe set context_switch and wait till they all pile in,
991 * then have them wait on a GC condition variable.
993 IF_DEBUG(scheduler,belch("--<< thread %ld (%p; %s) stopped: HeapOverflow",
994 t->id, t, whatNext_strs[t->what_next]));
997 ASSERT(!is_on_queue(t,CurrentProc));
999 /* Currently we emit a DESCHEDULE event before GC in GUM.
1000 ToDo: either add separate event to distinguish SYSTEM time from rest
1001 or just nuke this DESCHEDULE (and the following SCHEDULE) */
1002 if (0 && RtsFlags.ParFlags.ParStats.Full) {
1003 DumpRawGranEvent(CURRENT_PROC, CURRENT_PROC,
1004 GR_DESCHEDULE, t, (StgClosure *)NULL, 0, 0);
1005 emitSchedule = rtsTrue;
1009 ready_to_gc = rtsTrue;
1010 context_switch = 1; /* stop other threads ASAP */
1011 PUSH_ON_RUN_QUEUE(t);
1012 /* actual GC is done at the end of the while loop */
1018 DumpGranEvent(GR_DESCHEDULE, t));
1019 globalGranStats.tot_stackover++;
1022 // DumpGranEvent(GR_DESCHEDULE, t);
1023 globalParStats.tot_stackover++;
1025 IF_DEBUG(scheduler,belch("--<< thread %ld (%p; %s) stopped, StackOverflow",
1026 t->id, t, whatNext_strs[t->what_next]));
1027 /* just adjust the stack for this thread, then pop it back
1033 /* enlarge the stack */
1034 StgTSO *new_t = threadStackOverflow(t);
1036 /* This TSO has moved, so update any pointers to it from the
1037 * main thread stack. It better not be on any other queues...
1038 * (it shouldn't be).
1040 for (m = main_threads; m != NULL; m = m->link) {
1045 threadPaused(new_t);
1046 PUSH_ON_RUN_QUEUE(new_t);
1050 case ThreadYielding:
1053 DumpGranEvent(GR_DESCHEDULE, t));
1054 globalGranStats.tot_yields++;
1057 // DumpGranEvent(GR_DESCHEDULE, t);
1058 globalParStats.tot_yields++;
1060 /* put the thread back on the run queue. Then, if we're ready to
1061 * GC, check whether this is the last task to stop. If so, wake
1062 * up the GC thread. getThread will block during a GC until the
1066 if (t->what_next == ThreadEnterInterp) {
1067 /* ToDo: or maybe a timer expired when we were in Hugs?
1068 * or maybe someone hit ctrl-C
1070 belch("--<< thread %ld (%p; %s) stopped to switch to Hugs",
1071 t->id, t, whatNext_strs[t->what_next]);
1073 belch("--<< thread %ld (%p; %s) stopped, yielding",
1074 t->id, t, whatNext_strs[t->what_next]);
1081 //belch("&& Doing sanity check on yielding TSO %ld.", t->id);
1083 ASSERT(t->link == END_TSO_QUEUE);
1085 ASSERT(!is_on_queue(t,CurrentProc));
1088 //belch("&& Doing sanity check on all ThreadQueues (and their TSOs).");
1089 checkThreadQsSanity(rtsTrue));
1092 if (RtsFlags.ParFlags.doFairScheduling) {
1093 /* this does round-robin scheduling; good for concurrency */
1094 APPEND_TO_RUN_QUEUE(t);
1096 /* this does unfair scheduling; good for parallelism */
1097 PUSH_ON_RUN_QUEUE(t);
1100 /* this does round-robin scheduling; good for concurrency */
1101 APPEND_TO_RUN_QUEUE(t);
1104 /* add a ContinueThread event to actually process the thread */
1105 new_event(CurrentProc, CurrentProc, CurrentTime[CurrentProc],
1107 t, (StgClosure*)NULL, (rtsSpark*)NULL);
1109 belch("GRAN: eventq and runnableq after adding yielded thread to queue again:");
1118 belch("--<< thread %ld (%p; %s) stopped, blocking on node %p [PE %d] with BQ: ",
1119 t->id, t, whatNext_strs[t->what_next], t->block_info.closure, (t->block_info.closure==(StgClosure*)NULL ? 99 : where_is(t->block_info.closure)));
1120 if (t->block_info.closure!=(StgClosure*)NULL) print_bq(t->block_info.closure));
1122 // ??? needed; should emit block before
1124 DumpGranEvent(GR_DESCHEDULE, t));
1125 prune_eventq(t, (StgClosure *)NULL); // prune ContinueThreads for t
1128 ASSERT(procStatus[CurrentProc]==Busy ||
1129 ((procStatus[CurrentProc]==Fetching) &&
1130 (t->block_info.closure!=(StgClosure*)NULL)));
1131 if (run_queue_hds[CurrentProc] == END_TSO_QUEUE &&
1132 !(!RtsFlags.GranFlags.DoAsyncFetch &&
1133 procStatus[CurrentProc]==Fetching))
1134 procStatus[CurrentProc] = Idle;
1138 belch("--<< thread %ld (%p; %s) stopped, blocking on node %p with BQ: ",
1139 t->id, t, whatNext_strs[t->what_next], t->block_info.closure));
1142 if (t->block_info.closure!=(StgClosure*)NULL)
1143 print_bq(t->block_info.closure));
1145 /* Send a fetch (if BlockedOnGA) and dump event to log file */
1148 /* whatever we schedule next, we must log that schedule */
1149 emitSchedule = rtsTrue;
1152 /* don't need to do anything. Either the thread is blocked on
1153 * I/O, in which case we'll have called addToBlockedQueue
1154 * previously, or it's blocked on an MVar or Blackhole, in which
1155 * case it'll be on the relevant queue already.
1158 fprintf(stderr, "--<< thread %d (%p) stopped: ", t->id, t);
1159 printThreadBlockage(t);
1160 fprintf(stderr, "\n"));
1162 /* Only for dumping event to log file
1163 ToDo: do I need this in GranSim, too?
1170 case ThreadFinished:
1171 /* Need to check whether this was a main thread, and if so, signal
1172 * the task that started it with the return value. If we have no
1173 * more main threads, we probably need to stop all the tasks until
1176 /* We also end up here if the thread kills itself with an
1177 * uncaught exception, see Exception.hc.
1179 IF_DEBUG(scheduler,belch("--++ thread %d (%p) finished", t->id, t));
1181 endThread(t, CurrentProc); // clean-up the thread
1183 /* For now all are advisory -- HWL */
1184 //if(t->priority==AdvisoryPriority) ??
1185 advisory_thread_count--;
1188 if(t->dist.priority==RevalPriority)
1192 if (RtsFlags.ParFlags.ParStats.Full &&
1193 !RtsFlags.ParFlags.ParStats.Suppressed)
1194 DumpEndEvent(CURRENT_PROC, t, rtsFalse /* not mandatory */);
1199 barf("schedule: invalid thread return code %d", (int)ret);
1203 cap->link = free_capabilities;
1204 free_capabilities = cap;
1205 n_free_capabilities++;
1209 if (ready_to_gc && n_free_capabilities == RtsFlags.ParFlags.nNodes)
1214 /* everybody back, start the GC.
1215 * Could do it in this thread, or signal a condition var
1216 * to do it in another thread. Either way, we need to
1217 * broadcast on gc_pending_cond afterward.
1220 IF_DEBUG(scheduler,sched_belch("doing GC"));
1222 GarbageCollect(GetRoots,rtsFalse);
1223 ready_to_gc = rtsFalse;
1225 pthread_cond_broadcast(&gc_pending_cond);
1228 /* add a ContinueThread event to continue execution of current thread */
1229 new_event(CurrentProc, CurrentProc, CurrentTime[CurrentProc],
1231 t, (StgClosure*)NULL, (rtsSpark*)NULL);
1233 fprintf(stderr, "GRAN: eventq and runnableq after Garbage collection:\n");
1240 IF_GRAN_DEBUG(unused,
1241 print_eventq(EventHd));
1243 event = get_next_event();
1247 /* ToDo: wait for next message to arrive rather than busy wait */
1252 t = take_off_run_queue(END_TSO_QUEUE);
1255 } /* end of while(1) */
1256 IF_PAR_DEBUG(verbose,
1257 belch("== Leaving schedule() after having received Finish"));
1260 /* ---------------------------------------------------------------------------
1261 * deleteAllThreads(): kill all the live threads.
1263 * This is used when we catch a user interrupt (^C), before performing
1264 * any necessary cleanups and running finalizers.
1265 * ------------------------------------------------------------------------- */
1267 void deleteAllThreads ( void )
1270 IF_DEBUG(scheduler,sched_belch("deleting all threads"));
1271 for (t = run_queue_hd; t != END_TSO_QUEUE; t = t->link) {
1274 for (t = blocked_queue_hd; t != END_TSO_QUEUE; t = t->link) {
1277 for (t = sleeping_queue; t != END_TSO_QUEUE; t = t->link) {
1280 run_queue_hd = run_queue_tl = END_TSO_QUEUE;
1281 blocked_queue_hd = blocked_queue_tl = END_TSO_QUEUE;
1282 sleeping_queue = END_TSO_QUEUE;
1285 /* startThread and insertThread are now in GranSim.c -- HWL */
1287 //@node Suspend and Resume, Run queue code, Main scheduling loop, Main scheduling code
1288 //@subsection Suspend and Resume
1290 /* ---------------------------------------------------------------------------
1291 * Suspending & resuming Haskell threads.
1293 * When making a "safe" call to C (aka _ccall_GC), the task gives back
1294 * its capability before calling the C function. This allows another
1295 * task to pick up the capability and carry on running Haskell
1296 * threads. It also means that if the C call blocks, it won't lock
1299 * The Haskell thread making the C call is put to sleep for the
1300 * duration of the call, on the susepended_ccalling_threads queue. We
1301 * give out a token to the task, which it can use to resume the thread
1302 * on return from the C function.
1303 * ------------------------------------------------------------------------- */
1306 suspendThread( Capability *cap )
1310 ACQUIRE_LOCK(&sched_mutex);
1313 sched_belch("thread %d did a _ccall_gc", cap->rCurrentTSO->id));
1315 threadPaused(cap->rCurrentTSO);
1316 cap->rCurrentTSO->link = suspended_ccalling_threads;
1317 suspended_ccalling_threads = cap->rCurrentTSO;
1319 /* Use the thread ID as the token; it should be unique */
1320 tok = cap->rCurrentTSO->id;
1323 cap->link = free_capabilities;
1324 free_capabilities = cap;
1325 n_free_capabilities++;
1328 RELEASE_LOCK(&sched_mutex);
1333 resumeThread( StgInt tok )
1335 StgTSO *tso, **prev;
1338 ACQUIRE_LOCK(&sched_mutex);
1340 prev = &suspended_ccalling_threads;
1341 for (tso = suspended_ccalling_threads;
1342 tso != END_TSO_QUEUE;
1343 prev = &tso->link, tso = tso->link) {
1344 if (tso->id == (StgThreadID)tok) {
1349 if (tso == END_TSO_QUEUE) {
1350 barf("resumeThread: thread not found");
1352 tso->link = END_TSO_QUEUE;
1355 while (free_capabilities == NULL) {
1356 IF_DEBUG(scheduler, sched_belch("waiting to resume"));
1357 pthread_cond_wait(&thread_ready_cond, &sched_mutex);
1358 IF_DEBUG(scheduler, sched_belch("resuming thread %d", tso->id));
1360 cap = free_capabilities;
1361 free_capabilities = cap->link;
1362 n_free_capabilities--;
1364 cap = &MainRegTable;
1367 cap->rCurrentTSO = tso;
1369 RELEASE_LOCK(&sched_mutex);
1374 /* ---------------------------------------------------------------------------
1376 * ------------------------------------------------------------------------ */
1377 static void unblockThread(StgTSO *tso);
1379 /* ---------------------------------------------------------------------------
1380 * Comparing Thread ids.
1382 * This is used from STG land in the implementation of the
1383 * instances of Eq/Ord for ThreadIds.
1384 * ------------------------------------------------------------------------ */
1386 int cmp_thread(const StgTSO *tso1, const StgTSO *tso2)
1388 StgThreadID id1 = tso1->id;
1389 StgThreadID id2 = tso2->id;
1391 if (id1 < id2) return (-1);
1392 if (id1 > id2) return 1;
1396 /* ---------------------------------------------------------------------------
1397 Create a new thread.
1399 The new thread starts with the given stack size. Before the
1400 scheduler can run, however, this thread needs to have a closure
1401 (and possibly some arguments) pushed on its stack. See
1402 pushClosure() in Schedule.h.
1404 createGenThread() and createIOThread() (in SchedAPI.h) are
1405 convenient packaged versions of this function.
1407 currently pri (priority) is only used in a GRAN setup -- HWL
1408 ------------------------------------------------------------------------ */
1409 //@cindex createThread
1411 /* currently pri (priority) is only used in a GRAN setup -- HWL */
1413 createThread(nat stack_size, StgInt pri)
1415 return createThread_(stack_size, rtsFalse, pri);
1419 createThread_(nat size, rtsBool have_lock, StgInt pri)
1423 createThread(nat stack_size)
1425 return createThread_(stack_size, rtsFalse);
1429 createThread_(nat size, rtsBool have_lock)
1436 /* First check whether we should create a thread at all */
1438 /* check that no more than RtsFlags.ParFlags.maxThreads threads are created */
1439 if (advisory_thread_count >= RtsFlags.ParFlags.maxThreads) {
1441 belch("{createThread}Daq ghuH: refusing to create another thread; no more than %d threads allowed (currently %d)",
1442 RtsFlags.ParFlags.maxThreads, advisory_thread_count);
1443 return END_TSO_QUEUE;
1449 ASSERT(!RtsFlags.GranFlags.Light || CurrentProc==0);
1452 // ToDo: check whether size = stack_size - TSO_STRUCT_SIZEW
1454 /* catch ridiculously small stack sizes */
1455 if (size < MIN_STACK_WORDS + TSO_STRUCT_SIZEW) {
1456 size = MIN_STACK_WORDS + TSO_STRUCT_SIZEW;
1459 stack_size = size - TSO_STRUCT_SIZEW;
1461 tso = (StgTSO *)allocate(size);
1462 TICK_ALLOC_TSO(size-TSO_STRUCT_SIZEW, 0);
1464 SET_HDR(tso, &stg_TSO_info, CCS_SYSTEM);
1466 SET_GRAN_HDR(tso, ThisPE);
1468 tso->what_next = ThreadEnterGHC;
1470 /* tso->id needs to be unique. For now we use a heavyweight mutex to
1471 * protect the increment operation on next_thread_id.
1472 * In future, we could use an atomic increment instead.
1474 if (!have_lock) { ACQUIRE_LOCK(&sched_mutex); }
1475 tso->id = next_thread_id++;
1476 if (!have_lock) { RELEASE_LOCK(&sched_mutex); }
1478 tso->why_blocked = NotBlocked;
1479 tso->blocked_exceptions = NULL;
1481 tso->stack_size = stack_size;
1482 tso->max_stack_size = round_to_mblocks(RtsFlags.GcFlags.maxStkSize)
1484 tso->sp = (P_)&(tso->stack) + stack_size;
1487 tso->prof.CCCS = CCS_MAIN;
1490 /* put a stop frame on the stack */
1491 tso->sp -= sizeofW(StgStopFrame);
1492 SET_HDR((StgClosure*)tso->sp,(StgInfoTable *)&stg_stop_thread_info,CCS_SYSTEM);
1493 tso->su = (StgUpdateFrame*)tso->sp;
1497 tso->link = END_TSO_QUEUE;
1498 /* uses more flexible routine in GranSim */
1499 insertThread(tso, CurrentProc);
1501 /* In a non-GranSim setup the pushing of a TSO onto the runq is separated
1507 if (RtsFlags.GranFlags.GranSimStats.Full)
1508 DumpGranEvent(GR_START,tso);
1510 if (RtsFlags.ParFlags.ParStats.Full)
1511 DumpGranEvent(GR_STARTQ,tso);
1512 /* HACk to avoid SCHEDULE
1516 /* Link the new thread on the global thread list.
1518 tso->global_link = all_threads;
1522 tso->dist.priority = MandatoryPriority; //by default that is...
1526 tso->gran.pri = pri;
1528 tso->gran.magic = TSO_MAGIC; // debugging only
1530 tso->gran.sparkname = 0;
1531 tso->gran.startedat = CURRENT_TIME;
1532 tso->gran.exported = 0;
1533 tso->gran.basicblocks = 0;
1534 tso->gran.allocs = 0;
1535 tso->gran.exectime = 0;
1536 tso->gran.fetchtime = 0;
1537 tso->gran.fetchcount = 0;
1538 tso->gran.blocktime = 0;
1539 tso->gran.blockcount = 0;
1540 tso->gran.blockedat = 0;
1541 tso->gran.globalsparks = 0;
1542 tso->gran.localsparks = 0;
1543 if (RtsFlags.GranFlags.Light)
1544 tso->gran.clock = Now; /* local clock */
1546 tso->gran.clock = 0;
1548 IF_DEBUG(gran,printTSO(tso));
1551 tso->par.magic = TSO_MAGIC; // debugging only
1553 tso->par.sparkname = 0;
1554 tso->par.startedat = CURRENT_TIME;
1555 tso->par.exported = 0;
1556 tso->par.basicblocks = 0;
1557 tso->par.allocs = 0;
1558 tso->par.exectime = 0;
1559 tso->par.fetchtime = 0;
1560 tso->par.fetchcount = 0;
1561 tso->par.blocktime = 0;
1562 tso->par.blockcount = 0;
1563 tso->par.blockedat = 0;
1564 tso->par.globalsparks = 0;
1565 tso->par.localsparks = 0;
1569 globalGranStats.tot_threads_created++;
1570 globalGranStats.threads_created_on_PE[CurrentProc]++;
1571 globalGranStats.tot_sq_len += spark_queue_len(CurrentProc);
1572 globalGranStats.tot_sq_probes++;
1574 // collect parallel global statistics (currently done together with GC stats)
1575 if (RtsFlags.ParFlags.ParStats.Global &&
1576 RtsFlags.GcFlags.giveStats > NO_GC_STATS) {
1577 //fprintf(stderr, "Creating thread %d @ %11.2f\n", tso->id, usertime());
1578 globalParStats.tot_threads_created++;
1584 belch("==__ schedule: Created TSO %d (%p);",
1585 CurrentProc, tso, tso->id));
1587 IF_PAR_DEBUG(verbose,
1588 belch("==__ schedule: Created TSO %d (%p); %d threads active",
1589 tso->id, tso, advisory_thread_count));
1591 IF_DEBUG(scheduler,sched_belch("created thread %ld, stack size = %lx words",
1592 tso->id, tso->stack_size));
1599 all parallel thread creation calls should fall through the following routine.
1602 createSparkThread(rtsSpark spark)
1604 ASSERT(spark != (rtsSpark)NULL);
1605 if (advisory_thread_count >= RtsFlags.ParFlags.maxThreads)
1607 barf("{createSparkThread}Daq ghuH: refusing to create another thread; no more than %d threads allowed (currently %d)",
1608 RtsFlags.ParFlags.maxThreads, advisory_thread_count);
1609 return END_TSO_QUEUE;
1613 tso = createThread_(RtsFlags.GcFlags.initialStkSize, rtsTrue);
1614 if (tso==END_TSO_QUEUE)
1615 barf("createSparkThread: Cannot create TSO");
1617 tso->priority = AdvisoryPriority;
1619 pushClosure(tso,spark);
1620 PUSH_ON_RUN_QUEUE(tso);
1621 advisory_thread_count++;
1628 Turn a spark into a thread.
1629 ToDo: fix for SMP (needs to acquire SCHED_MUTEX!)
1632 //@cindex activateSpark
1634 activateSpark (rtsSpark spark)
1638 tso = createSparkThread(spark);
1639 if (RtsFlags.ParFlags.ParStats.Full) {
1640 //ASSERT(run_queue_hd == END_TSO_QUEUE); // I think ...
1641 IF_PAR_DEBUG(verbose,
1642 belch("==^^ activateSpark: turning spark of closure %p (%s) into a thread",
1643 (StgClosure *)spark, info_type((StgClosure *)spark)));
1645 // ToDo: fwd info on local/global spark to thread -- HWL
1646 // tso->gran.exported = spark->exported;
1647 // tso->gran.locked = !spark->global;
1648 // tso->gran.sparkname = spark->name;
1654 /* ---------------------------------------------------------------------------
1657 * scheduleThread puts a thread on the head of the runnable queue.
1658 * This will usually be done immediately after a thread is created.
1659 * The caller of scheduleThread must create the thread using e.g.
1660 * createThread and push an appropriate closure
1661 * on this thread's stack before the scheduler is invoked.
1662 * ------------------------------------------------------------------------ */
1665 scheduleThread(StgTSO *tso)
1667 if (tso==END_TSO_QUEUE){
1672 ACQUIRE_LOCK(&sched_mutex);
1674 /* Put the new thread on the head of the runnable queue. The caller
1675 * better push an appropriate closure on this thread's stack
1676 * beforehand. In the SMP case, the thread may start running as
1677 * soon as we release the scheduler lock below.
1679 PUSH_ON_RUN_QUEUE(tso);
1683 IF_DEBUG(scheduler,printTSO(tso));
1685 RELEASE_LOCK(&sched_mutex);
1688 /* ---------------------------------------------------------------------------
1691 * Start up Posix threads to run each of the scheduler tasks.
1692 * I believe the task ids are not needed in the system as defined.
1694 * ------------------------------------------------------------------------ */
1696 #if defined(PAR) || defined(SMP)
1698 taskStart(void) /* ( void *arg STG_UNUSED) */
1700 scheduleThread(END_TSO_QUEUE);
1704 /* ---------------------------------------------------------------------------
1707 * Initialise the scheduler. This resets all the queues - if the
1708 * queues contained any threads, they'll be garbage collected at the
1711 * This now calls startTasks(), so should only be called once! KH @ 25/10/99
1712 * ------------------------------------------------------------------------ */
1716 term_handler(int sig STG_UNUSED)
1719 ACQUIRE_LOCK(&term_mutex);
1721 RELEASE_LOCK(&term_mutex);
1726 //@cindex initScheduler
1733 for (i=0; i<=MAX_PROC; i++) {
1734 run_queue_hds[i] = END_TSO_QUEUE;
1735 run_queue_tls[i] = END_TSO_QUEUE;
1736 blocked_queue_hds[i] = END_TSO_QUEUE;
1737 blocked_queue_tls[i] = END_TSO_QUEUE;
1738 ccalling_threadss[i] = END_TSO_QUEUE;
1739 sleeping_queue = END_TSO_QUEUE;
1742 run_queue_hd = END_TSO_QUEUE;
1743 run_queue_tl = END_TSO_QUEUE;
1744 blocked_queue_hd = END_TSO_QUEUE;
1745 blocked_queue_tl = END_TSO_QUEUE;
1746 sleeping_queue = END_TSO_QUEUE;
1749 suspended_ccalling_threads = END_TSO_QUEUE;
1751 main_threads = NULL;
1752 all_threads = END_TSO_QUEUE;
1757 RtsFlags.ConcFlags.ctxtSwitchTicks =
1758 RtsFlags.ConcFlags.ctxtSwitchTime / TICK_MILLISECS;
1760 /* Install the SIGHUP handler */
1763 struct sigaction action,oact;
1765 action.sa_handler = term_handler;
1766 sigemptyset(&action.sa_mask);
1767 action.sa_flags = 0;
1768 if (sigaction(SIGTERM, &action, &oact) != 0) {
1769 barf("can't install TERM handler");
1775 /* Allocate N Capabilities */
1778 Capability *cap, *prev;
1781 for (i = 0; i < RtsFlags.ParFlags.nNodes; i++) {
1782 cap = stgMallocBytes(sizeof(Capability), "initScheduler:capabilities");
1786 free_capabilities = cap;
1787 n_free_capabilities = RtsFlags.ParFlags.nNodes;
1789 IF_DEBUG(scheduler,fprintf(stderr,"scheduler: Allocated %d capabilities\n",
1790 n_free_capabilities););
1793 #if defined(SMP) || defined(PAR)
1806 /* make some space for saving all the thread ids */
1807 task_ids = stgMallocBytes(RtsFlags.ParFlags.nNodes * sizeof(task_info),
1808 "initScheduler:task_ids");
1810 /* and create all the threads */
1811 for (i = 0; i < RtsFlags.ParFlags.nNodes; i++) {
1812 r = pthread_create(&tid,NULL,taskStart,NULL);
1814 barf("startTasks: Can't create new Posix thread");
1816 task_ids[i].id = tid;
1817 task_ids[i].mut_time = 0.0;
1818 task_ids[i].mut_etime = 0.0;
1819 task_ids[i].gc_time = 0.0;
1820 task_ids[i].gc_etime = 0.0;
1821 task_ids[i].elapsedtimestart = elapsedtime();
1822 IF_DEBUG(scheduler,fprintf(stderr,"scheduler: Started task: %ld\n",tid););
1828 exitScheduler( void )
1833 /* Don't want to use pthread_cancel, since we'd have to install
1834 * these silly exception handlers (pthread_cleanup_{push,pop}) around
1838 /* Cancel all our tasks */
1839 for (i = 0; i < RtsFlags.ParFlags.nNodes; i++) {
1840 pthread_cancel(task_ids[i].id);
1843 /* Wait for all the tasks to terminate */
1844 for (i = 0; i < RtsFlags.ParFlags.nNodes; i++) {
1845 IF_DEBUG(scheduler,fprintf(stderr,"scheduler: waiting for task %ld\n",
1847 pthread_join(task_ids[i].id, NULL);
1851 /* Send 'em all a SIGHUP. That should shut 'em up.
1853 await_death = RtsFlags.ParFlags.nNodes;
1854 for (i = 0; i < RtsFlags.ParFlags.nNodes; i++) {
1855 pthread_kill(task_ids[i].id,SIGTERM);
1857 while (await_death > 0) {
1863 /* -----------------------------------------------------------------------------
1864 Managing the per-task allocation areas.
1866 Each capability comes with an allocation area. These are
1867 fixed-length block lists into which allocation can be done.
1869 ToDo: no support for two-space collection at the moment???
1870 -------------------------------------------------------------------------- */
1872 /* -----------------------------------------------------------------------------
1873 * waitThread is the external interface for running a new computation
1874 * and waiting for the result.
1876 * In the non-SMP case, we create a new main thread, push it on the
1877 * main-thread stack, and invoke the scheduler to run it. The
1878 * scheduler will return when the top main thread on the stack has
1879 * completed or died, and fill in the necessary fields of the
1880 * main_thread structure.
1882 * In the SMP case, we create a main thread as before, but we then
1883 * create a new condition variable and sleep on it. When our new
1884 * main thread has completed, we'll be woken up and the status/result
1885 * will be in the main_thread struct.
1886 * -------------------------------------------------------------------------- */
1889 howManyThreadsAvail ( void )
1893 for (q = run_queue_hd; q != END_TSO_QUEUE; q = q->link)
1895 for (q = blocked_queue_hd; q != END_TSO_QUEUE; q = q->link)
1897 for (q = sleeping_queue; q != END_TSO_QUEUE; q = q->link)
1903 finishAllThreads ( void )
1906 while (run_queue_hd != END_TSO_QUEUE) {
1907 waitThread ( run_queue_hd, NULL );
1909 while (blocked_queue_hd != END_TSO_QUEUE) {
1910 waitThread ( blocked_queue_hd, NULL );
1912 while (sleeping_queue != END_TSO_QUEUE) {
1913 waitThread ( blocked_queue_hd, NULL );
1916 (blocked_queue_hd != END_TSO_QUEUE ||
1917 run_queue_hd != END_TSO_QUEUE ||
1918 sleeping_queue != END_TSO_QUEUE);
1922 waitThread(StgTSO *tso, /*out*/StgClosure **ret)
1925 SchedulerStatus stat;
1927 ACQUIRE_LOCK(&sched_mutex);
1929 m = stgMallocBytes(sizeof(StgMainThread), "waitThread");
1935 pthread_cond_init(&m->wakeup, NULL);
1938 m->link = main_threads;
1941 IF_DEBUG(scheduler, fprintf(stderr, "== scheduler: new main thread (%d)\n",
1946 pthread_cond_wait(&m->wakeup, &sched_mutex);
1947 } while (m->stat == NoStatus);
1949 /* GranSim specific init */
1950 CurrentTSO = m->tso; // the TSO to run
1951 procStatus[MainProc] = Busy; // status of main PE
1952 CurrentProc = MainProc; // PE to run it on
1957 ASSERT(m->stat != NoStatus);
1963 pthread_cond_destroy(&m->wakeup);
1966 IF_DEBUG(scheduler, fprintf(stderr, "== scheduler: main thread (%d) finished\n",
1970 RELEASE_LOCK(&sched_mutex);
1975 //@node Run queue code, Garbage Collextion Routines, Suspend and Resume, Main scheduling code
1976 //@subsection Run queue code
1980 NB: In GranSim we have many run queues; run_queue_hd is actually a macro
1981 unfolding to run_queue_hds[CurrentProc], thus CurrentProc is an
1982 implicit global variable that has to be correct when calling these
1986 /* Put the new thread on the head of the runnable queue.
1987 * The caller of createThread better push an appropriate closure
1988 * on this thread's stack before the scheduler is invoked.
1990 static /* inline */ void
1991 add_to_run_queue(tso)
1994 ASSERT(tso!=run_queue_hd && tso!=run_queue_tl);
1995 tso->link = run_queue_hd;
1997 if (run_queue_tl == END_TSO_QUEUE) {
2002 /* Put the new thread at the end of the runnable queue. */
2003 static /* inline */ void
2004 push_on_run_queue(tso)
2007 ASSERT(get_itbl((StgClosure *)tso)->type == TSO);
2008 ASSERT(run_queue_hd!=NULL && run_queue_tl!=NULL);
2009 ASSERT(tso!=run_queue_hd && tso!=run_queue_tl);
2010 if (run_queue_hd == END_TSO_QUEUE) {
2013 run_queue_tl->link = tso;
2019 Should be inlined because it's used very often in schedule. The tso
2020 argument is actually only needed in GranSim, where we want to have the
2021 possibility to schedule *any* TSO on the run queue, irrespective of the
2022 actual ordering. Therefore, if tso is not the nil TSO then we traverse
2023 the run queue and dequeue the tso, adjusting the links in the queue.
2025 //@cindex take_off_run_queue
2026 static /* inline */ StgTSO*
2027 take_off_run_queue(StgTSO *tso) {
2031 qetlaHbogh Qu' ngaSbogh ghomDaQ {tso} yIteq!
2033 if tso is specified, unlink that tso from the run_queue (doesn't have
2034 to be at the beginning of the queue); GranSim only
2036 if (tso!=END_TSO_QUEUE) {
2037 /* find tso in queue */
2038 for (t=run_queue_hd, prev=END_TSO_QUEUE;
2039 t!=END_TSO_QUEUE && t!=tso;
2043 /* now actually dequeue the tso */
2044 if (prev!=END_TSO_QUEUE) {
2045 ASSERT(run_queue_hd!=t);
2046 prev->link = t->link;
2048 /* t is at beginning of thread queue */
2049 ASSERT(run_queue_hd==t);
2050 run_queue_hd = t->link;
2052 /* t is at end of thread queue */
2053 if (t->link==END_TSO_QUEUE) {
2054 ASSERT(t==run_queue_tl);
2055 run_queue_tl = prev;
2057 ASSERT(run_queue_tl!=t);
2059 t->link = END_TSO_QUEUE;
2061 /* take tso from the beginning of the queue; std concurrent code */
2063 if (t != END_TSO_QUEUE) {
2064 run_queue_hd = t->link;
2065 t->link = END_TSO_QUEUE;
2066 if (run_queue_hd == END_TSO_QUEUE) {
2067 run_queue_tl = END_TSO_QUEUE;
2076 //@node Garbage Collextion Routines, Blocking Queue Routines, Run queue code, Main scheduling code
2077 //@subsection Garbage Collextion Routines
2079 /* ---------------------------------------------------------------------------
2080 Where are the roots that we know about?
2082 - all the threads on the runnable queue
2083 - all the threads on the blocked queue
2084 - all the threads on the sleeping queue
2085 - all the thread currently executing a _ccall_GC
2086 - all the "main threads"
2088 ------------------------------------------------------------------------ */
2090 /* This has to be protected either by the scheduler monitor, or by the
2091 garbage collection monitor (probably the latter).
2095 static void GetRoots(evac_fn evac)
2102 for (i=0; i<=RtsFlags.GranFlags.proc; i++) {
2103 if ((run_queue_hds[i] != END_TSO_QUEUE) && ((run_queue_hds[i] != NULL)))
2104 evac((StgClosure **)&run_queue_hds[i]);
2105 if ((run_queue_tls[i] != END_TSO_QUEUE) && ((run_queue_tls[i] != NULL)))
2106 evac((StgClosure **)&run_queue_tls[i]);
2108 if ((blocked_queue_hds[i] != END_TSO_QUEUE) && ((blocked_queue_hds[i] != NULL)))
2109 evac((StgClosure **)&blocked_queue_hds[i]);
2110 if ((blocked_queue_tls[i] != END_TSO_QUEUE) && ((blocked_queue_tls[i] != NULL)))
2111 evac((StgClosure **)&blocked_queue_tls[i]);
2112 if ((ccalling_threadss[i] != END_TSO_QUEUE) && ((ccalling_threadss[i] != NULL)))
2113 evac((StgClosure **)&ccalling_threads[i]);
2120 if (run_queue_hd != END_TSO_QUEUE) {
2121 ASSERT(run_queue_tl != END_TSO_QUEUE);
2122 evac((StgClosure **)&run_queue_hd);
2123 evac((StgClosure **)&run_queue_tl);
2126 if (blocked_queue_hd != END_TSO_QUEUE) {
2127 ASSERT(blocked_queue_tl != END_TSO_QUEUE);
2128 evac((StgClosure **)&blocked_queue_hd);
2129 evac((StgClosure **)&blocked_queue_tl);
2132 if (sleeping_queue != END_TSO_QUEUE) {
2133 evac((StgClosure **)&sleeping_queue);
2137 for (m = main_threads; m != NULL; m = m->link) {
2138 evac((StgClosure **)&m->tso);
2140 if (suspended_ccalling_threads != END_TSO_QUEUE) {
2141 evac((StgClosure **)&suspended_ccalling_threads);
2144 #if defined(SMP) || defined(PAR) || defined(GRAN)
2145 markSparkQueue(evac);
2149 /* -----------------------------------------------------------------------------
2152 This is the interface to the garbage collector from Haskell land.
2153 We provide this so that external C code can allocate and garbage
2154 collect when called from Haskell via _ccall_GC.
2156 It might be useful to provide an interface whereby the programmer
2157 can specify more roots (ToDo).
2159 This needs to be protected by the GC condition variable above. KH.
2160 -------------------------------------------------------------------------- */
2162 void (*extra_roots)(evac_fn);
2167 GarbageCollect(GetRoots,rtsFalse);
2171 performMajorGC(void)
2173 GarbageCollect(GetRoots,rtsTrue);
2177 AllRoots(evac_fn evac)
2179 GetRoots(evac); // the scheduler's roots
2180 extra_roots(evac); // the user's roots
2184 performGCWithRoots(void (*get_roots)(evac_fn))
2186 extra_roots = get_roots;
2187 GarbageCollect(AllRoots,rtsFalse);
2190 /* -----------------------------------------------------------------------------
2193 If the thread has reached its maximum stack size, then raise the
2194 StackOverflow exception in the offending thread. Otherwise
2195 relocate the TSO into a larger chunk of memory and adjust its stack
2197 -------------------------------------------------------------------------- */
2200 threadStackOverflow(StgTSO *tso)
2202 nat new_stack_size, new_tso_size, diff, stack_words;
2206 IF_DEBUG(sanity,checkTSO(tso));
2207 if (tso->stack_size >= tso->max_stack_size) {
2210 belch("@@ threadStackOverflow of TSO %d (%p): stack too large (now %ld; max is %ld",
2211 tso->id, tso, tso->stack_size, tso->max_stack_size);
2212 /* If we're debugging, just print out the top of the stack */
2213 printStackChunk(tso->sp, stg_min(tso->stack+tso->stack_size,
2216 /* Send this thread the StackOverflow exception */
2217 raiseAsync(tso, (StgClosure *)stackOverflow_closure);
2221 /* Try to double the current stack size. If that takes us over the
2222 * maximum stack size for this thread, then use the maximum instead.
2223 * Finally round up so the TSO ends up as a whole number of blocks.
2225 new_stack_size = stg_min(tso->stack_size * 2, tso->max_stack_size);
2226 new_tso_size = (nat)BLOCK_ROUND_UP(new_stack_size * sizeof(W_) +
2227 TSO_STRUCT_SIZE)/sizeof(W_);
2228 new_tso_size = round_to_mblocks(new_tso_size); /* Be MBLOCK-friendly */
2229 new_stack_size = new_tso_size - TSO_STRUCT_SIZEW;
2231 IF_DEBUG(scheduler, fprintf(stderr,"== scheduler: increasing stack size from %d words to %d.\n", tso->stack_size, new_stack_size));
2233 dest = (StgTSO *)allocate(new_tso_size);
2234 TICK_ALLOC_TSO(new_tso_size-sizeofW(StgTSO),0);
2236 /* copy the TSO block and the old stack into the new area */
2237 memcpy(dest,tso,TSO_STRUCT_SIZE);
2238 stack_words = tso->stack + tso->stack_size - tso->sp;
2239 new_sp = (P_)dest + new_tso_size - stack_words;
2240 memcpy(new_sp, tso->sp, stack_words * sizeof(W_));
2242 /* relocate the stack pointers... */
2243 diff = (P_)new_sp - (P_)tso->sp; /* In *words* */
2244 dest->su = (StgUpdateFrame *) ((P_)dest->su + diff);
2246 dest->stack_size = new_stack_size;
2248 /* and relocate the update frame list */
2249 relocate_stack(dest, diff);
2251 /* Mark the old TSO as relocated. We have to check for relocated
2252 * TSOs in the garbage collector and any primops that deal with TSOs.
2254 * It's important to set the sp and su values to just beyond the end
2255 * of the stack, so we don't attempt to scavenge any part of the
2258 tso->what_next = ThreadRelocated;
2260 tso->sp = (P_)&(tso->stack[tso->stack_size]);
2261 tso->su = (StgUpdateFrame *)tso->sp;
2262 tso->why_blocked = NotBlocked;
2263 dest->mut_link = NULL;
2265 IF_PAR_DEBUG(verbose,
2266 belch("@@ threadStackOverflow of TSO %d (now at %p): stack size increased to %ld",
2267 tso->id, tso, tso->stack_size);
2268 /* If we're debugging, just print out the top of the stack */
2269 printStackChunk(tso->sp, stg_min(tso->stack+tso->stack_size,
2272 IF_DEBUG(sanity,checkTSO(tso));
2274 IF_DEBUG(scheduler,printTSO(dest));
2280 //@node Blocking Queue Routines, Exception Handling Routines, Garbage Collextion Routines, Main scheduling code
2281 //@subsection Blocking Queue Routines
2283 /* ---------------------------------------------------------------------------
2284 Wake up a queue that was blocked on some resource.
2285 ------------------------------------------------------------------------ */
2289 unblockCount ( StgBlockingQueueElement *bqe, StgClosure *node )
2294 unblockCount ( StgBlockingQueueElement *bqe, StgClosure *node )
2296 /* write RESUME events to log file and
2297 update blocked and fetch time (depending on type of the orig closure) */
2298 if (RtsFlags.ParFlags.ParStats.Full) {
2299 DumpRawGranEvent(CURRENT_PROC, CURRENT_PROC,
2300 GR_RESUMEQ, ((StgTSO *)bqe), ((StgTSO *)bqe)->block_info.closure,
2301 0, 0 /* spark_queue_len(ADVISORY_POOL) */);
2302 if (EMPTY_RUN_QUEUE())
2303 emitSchedule = rtsTrue;
2305 switch (get_itbl(node)->type) {
2307 ((StgTSO *)bqe)->par.fetchtime += CURRENT_TIME-((StgTSO *)bqe)->par.blockedat;
2312 ((StgTSO *)bqe)->par.blocktime += CURRENT_TIME-((StgTSO *)bqe)->par.blockedat;
2319 barf("{unblockOneLocked}Daq Qagh: unexpected closure in blocking queue");
2326 static StgBlockingQueueElement *
2327 unblockOneLocked(StgBlockingQueueElement *bqe, StgClosure *node)
2330 PEs node_loc, tso_loc;
2332 node_loc = where_is(node); // should be lifted out of loop
2333 tso = (StgTSO *)bqe; // wastes an assignment to get the type right
2334 tso_loc = where_is((StgClosure *)tso);
2335 if (IS_LOCAL_TO(PROCS(node),tso_loc)) { // TSO is local
2336 /* !fake_fetch => TSO is on CurrentProc is same as IS_LOCAL_TO */
2337 ASSERT(CurrentProc!=node_loc || tso_loc==CurrentProc);
2338 CurrentTime[CurrentProc] += RtsFlags.GranFlags.Costs.lunblocktime;
2339 // insertThread(tso, node_loc);
2340 new_event(tso_loc, tso_loc, CurrentTime[CurrentProc],
2342 tso, node, (rtsSpark*)NULL);
2343 tso->link = END_TSO_QUEUE; // overwrite link just to be sure
2346 } else { // TSO is remote (actually should be FMBQ)
2347 CurrentTime[CurrentProc] += RtsFlags.GranFlags.Costs.mpacktime +
2348 RtsFlags.GranFlags.Costs.gunblocktime +
2349 RtsFlags.GranFlags.Costs.latency;
2350 new_event(tso_loc, CurrentProc, CurrentTime[CurrentProc],
2352 tso, node, (rtsSpark*)NULL);
2353 tso->link = END_TSO_QUEUE; // overwrite link just to be sure
2356 /* the thread-queue-overhead is accounted for in either Resume or UnblockThread */
2358 fprintf(stderr," %s TSO %d (%p) [PE %d] (block_info.closure=%p) (next=%p) ,",
2359 (node_loc==tso_loc ? "Local" : "Global"),
2360 tso->id, tso, CurrentProc, tso->block_info.closure, tso->link));
2361 tso->block_info.closure = NULL;
2362 IF_DEBUG(scheduler,belch("-- Waking up thread %ld (%p)",
2366 static StgBlockingQueueElement *
2367 unblockOneLocked(StgBlockingQueueElement *bqe, StgClosure *node)
2369 StgBlockingQueueElement *next;
2371 switch (get_itbl(bqe)->type) {
2373 ASSERT(((StgTSO *)bqe)->why_blocked != NotBlocked);
2374 /* if it's a TSO just push it onto the run_queue */
2376 // ((StgTSO *)bqe)->link = END_TSO_QUEUE; // debugging?
2377 PUSH_ON_RUN_QUEUE((StgTSO *)bqe);
2379 unblockCount(bqe, node);
2380 /* reset blocking status after dumping event */
2381 ((StgTSO *)bqe)->why_blocked = NotBlocked;
2385 /* if it's a BLOCKED_FETCH put it on the PendingFetches list */
2387 bqe->link = (StgBlockingQueueElement *)PendingFetches;
2388 PendingFetches = (StgBlockedFetch *)bqe;
2392 /* can ignore this case in a non-debugging setup;
2393 see comments on RBHSave closures above */
2395 /* check that the closure is an RBHSave closure */
2396 ASSERT(get_itbl((StgClosure *)bqe) == &stg_RBH_Save_0_info ||
2397 get_itbl((StgClosure *)bqe) == &stg_RBH_Save_1_info ||
2398 get_itbl((StgClosure *)bqe) == &stg_RBH_Save_2_info);
2402 barf("{unblockOneLocked}Daq Qagh: Unexpected IP (%#lx; %s) in blocking queue at %#lx\n",
2403 get_itbl((StgClosure *)bqe), info_type((StgClosure *)bqe),
2407 IF_PAR_DEBUG(bq, fprintf(stderr, ", %p (%s)", bqe, info_type((StgClosure*)bqe)));
2411 #else /* !GRAN && !PAR */
2413 unblockOneLocked(StgTSO *tso)
2417 ASSERT(get_itbl(tso)->type == TSO);
2418 ASSERT(tso->why_blocked != NotBlocked);
2419 tso->why_blocked = NotBlocked;
2421 PUSH_ON_RUN_QUEUE(tso);
2423 IF_DEBUG(scheduler,sched_belch("waking up thread %ld", tso->id));
2428 #if defined(GRAN) || defined(PAR)
2429 inline StgBlockingQueueElement *
2430 unblockOne(StgBlockingQueueElement *bqe, StgClosure *node)
2432 ACQUIRE_LOCK(&sched_mutex);
2433 bqe = unblockOneLocked(bqe, node);
2434 RELEASE_LOCK(&sched_mutex);
2439 unblockOne(StgTSO *tso)
2441 ACQUIRE_LOCK(&sched_mutex);
2442 tso = unblockOneLocked(tso);
2443 RELEASE_LOCK(&sched_mutex);
2450 awakenBlockedQueue(StgBlockingQueueElement *q, StgClosure *node)
2452 StgBlockingQueueElement *bqe;
2457 belch("##-_ AwBQ for node %p on PE %d @ %ld by TSO %d (%p): ", \
2458 node, CurrentProc, CurrentTime[CurrentProc],
2459 CurrentTSO->id, CurrentTSO));
2461 node_loc = where_is(node);
2463 ASSERT(q == END_BQ_QUEUE ||
2464 get_itbl(q)->type == TSO || // q is either a TSO or an RBHSave
2465 get_itbl(q)->type == CONSTR); // closure (type constructor)
2466 ASSERT(is_unique(node));
2468 /* FAKE FETCH: magically copy the node to the tso's proc;
2469 no Fetch necessary because in reality the node should not have been
2470 moved to the other PE in the first place
2472 if (CurrentProc!=node_loc) {
2474 belch("## node %p is on PE %d but CurrentProc is %d (TSO %d); assuming fake fetch and adjusting bitmask (old: %#x)",
2475 node, node_loc, CurrentProc, CurrentTSO->id,
2476 // CurrentTSO, where_is(CurrentTSO),
2477 node->header.gran.procs));
2478 node->header.gran.procs = (node->header.gran.procs) | PE_NUMBER(CurrentProc);
2480 belch("## new bitmask of node %p is %#x",
2481 node, node->header.gran.procs));
2482 if (RtsFlags.GranFlags.GranSimStats.Global) {
2483 globalGranStats.tot_fake_fetches++;
2488 // ToDo: check: ASSERT(CurrentProc==node_loc);
2489 while (get_itbl(bqe)->type==TSO) { // q != END_TSO_QUEUE) {
2492 bqe points to the current element in the queue
2493 next points to the next element in the queue
2495 //tso = (StgTSO *)bqe; // wastes an assignment to get the type right
2496 //tso_loc = where_is(tso);
2498 bqe = unblockOneLocked(bqe, node);
2501 /* if this is the BQ of an RBH, we have to put back the info ripped out of
2502 the closure to make room for the anchor of the BQ */
2503 if (bqe!=END_BQ_QUEUE) {
2504 ASSERT(get_itbl(node)->type == RBH && get_itbl(bqe)->type == CONSTR);
2506 ASSERT((info_ptr==&RBH_Save_0_info) ||
2507 (info_ptr==&RBH_Save_1_info) ||
2508 (info_ptr==&RBH_Save_2_info));
2510 /* cf. convertToRBH in RBH.c for writing the RBHSave closure */
2511 ((StgRBH *)node)->blocking_queue = (StgBlockingQueueElement *)((StgRBHSave *)bqe)->payload[0];
2512 ((StgRBH *)node)->mut_link = (StgMutClosure *)((StgRBHSave *)bqe)->payload[1];
2515 belch("## Filled in RBH_Save for %p (%s) at end of AwBQ",
2516 node, info_type(node)));
2519 /* statistics gathering */
2520 if (RtsFlags.GranFlags.GranSimStats.Global) {
2521 // globalGranStats.tot_bq_processing_time += bq_processing_time;
2522 globalGranStats.tot_bq_len += len; // total length of all bqs awakened
2523 // globalGranStats.tot_bq_len_local += len_local; // same for local TSOs only
2524 globalGranStats.tot_awbq++; // total no. of bqs awakened
2527 fprintf(stderr,"## BQ Stats of %p: [%d entries] %s\n",
2528 node, len, (bqe!=END_BQ_QUEUE) ? "RBH" : ""));
2532 awakenBlockedQueue(StgBlockingQueueElement *q, StgClosure *node)
2534 StgBlockingQueueElement *bqe;
2536 ACQUIRE_LOCK(&sched_mutex);
2538 IF_PAR_DEBUG(verbose,
2539 belch("##-_ AwBQ for node %p on [%x]: ",
2543 if(get_itbl(q)->type == CONSTR || q==END_BQ_QUEUE) {
2544 IF_PAR_DEBUG(verbose, belch("## ... nothing to unblock so lets just return. RFP (BUG?)"));
2549 ASSERT(q == END_BQ_QUEUE ||
2550 get_itbl(q)->type == TSO ||
2551 get_itbl(q)->type == BLOCKED_FETCH ||
2552 get_itbl(q)->type == CONSTR);
2555 while (get_itbl(bqe)->type==TSO ||
2556 get_itbl(bqe)->type==BLOCKED_FETCH) {
2557 bqe = unblockOneLocked(bqe, node);
2559 RELEASE_LOCK(&sched_mutex);
2562 #else /* !GRAN && !PAR */
2564 awakenBlockedQueue(StgTSO *tso)
2566 ACQUIRE_LOCK(&sched_mutex);
2567 while (tso != END_TSO_QUEUE) {
2568 tso = unblockOneLocked(tso);
2570 RELEASE_LOCK(&sched_mutex);
2574 //@node Exception Handling Routines, Debugging Routines, Blocking Queue Routines, Main scheduling code
2575 //@subsection Exception Handling Routines
2577 /* ---------------------------------------------------------------------------
2579 - usually called inside a signal handler so it mustn't do anything fancy.
2580 ------------------------------------------------------------------------ */
2583 interruptStgRts(void)
2589 /* -----------------------------------------------------------------------------
2592 This is for use when we raise an exception in another thread, which
2594 This has nothing to do with the UnblockThread event in GranSim. -- HWL
2595 -------------------------------------------------------------------------- */
2597 #if defined(GRAN) || defined(PAR)
2599 NB: only the type of the blocking queue is different in GranSim and GUM
2600 the operations on the queue-elements are the same
2601 long live polymorphism!
2604 unblockThread(StgTSO *tso)
2606 StgBlockingQueueElement *t, **last;
2608 ACQUIRE_LOCK(&sched_mutex);
2609 switch (tso->why_blocked) {
2612 return; /* not blocked */
2615 ASSERT(get_itbl(tso->block_info.closure)->type == MVAR);
2617 StgBlockingQueueElement *last_tso = END_BQ_QUEUE;
2618 StgMVar *mvar = (StgMVar *)(tso->block_info.closure);
2620 last = (StgBlockingQueueElement **)&mvar->head;
2621 for (t = (StgBlockingQueueElement *)mvar->head;
2623 last = &t->link, last_tso = t, t = t->link) {
2624 if (t == (StgBlockingQueueElement *)tso) {
2625 *last = (StgBlockingQueueElement *)tso->link;
2626 if (mvar->tail == tso) {
2627 mvar->tail = (StgTSO *)last_tso;
2632 barf("unblockThread (MVAR): TSO not found");
2635 case BlockedOnBlackHole:
2636 ASSERT(get_itbl(tso->block_info.closure)->type == BLACKHOLE_BQ);
2638 StgBlockingQueue *bq = (StgBlockingQueue *)(tso->block_info.closure);
2640 last = &bq->blocking_queue;
2641 for (t = bq->blocking_queue;
2643 last = &t->link, t = t->link) {
2644 if (t == (StgBlockingQueueElement *)tso) {
2645 *last = (StgBlockingQueueElement *)tso->link;
2649 barf("unblockThread (BLACKHOLE): TSO not found");
2652 case BlockedOnException:
2654 StgTSO *target = tso->block_info.tso;
2656 ASSERT(get_itbl(target)->type == TSO);
2658 if (target->what_next == ThreadRelocated) {
2659 target = target->link;
2660 ASSERT(get_itbl(target)->type == TSO);
2663 ASSERT(target->blocked_exceptions != NULL);
2665 last = (StgBlockingQueueElement **)&target->blocked_exceptions;
2666 for (t = (StgBlockingQueueElement *)target->blocked_exceptions;
2668 last = &t->link, t = t->link) {
2669 ASSERT(get_itbl(t)->type == TSO);
2670 if (t == (StgBlockingQueueElement *)tso) {
2671 *last = (StgBlockingQueueElement *)tso->link;
2675 barf("unblockThread (Exception): TSO not found");
2679 case BlockedOnWrite:
2681 /* take TSO off blocked_queue */
2682 StgBlockingQueueElement *prev = NULL;
2683 for (t = (StgBlockingQueueElement *)blocked_queue_hd; t != END_BQ_QUEUE;
2684 prev = t, t = t->link) {
2685 if (t == (StgBlockingQueueElement *)tso) {
2687 blocked_queue_hd = (StgTSO *)t->link;
2688 if ((StgBlockingQueueElement *)blocked_queue_tl == t) {
2689 blocked_queue_tl = END_TSO_QUEUE;
2692 prev->link = t->link;
2693 if ((StgBlockingQueueElement *)blocked_queue_tl == t) {
2694 blocked_queue_tl = (StgTSO *)prev;
2700 barf("unblockThread (I/O): TSO not found");
2703 case BlockedOnDelay:
2705 /* take TSO off sleeping_queue */
2706 StgBlockingQueueElement *prev = NULL;
2707 for (t = (StgBlockingQueueElement *)sleeping_queue; t != END_BQ_QUEUE;
2708 prev = t, t = t->link) {
2709 if (t == (StgBlockingQueueElement *)tso) {
2711 sleeping_queue = (StgTSO *)t->link;
2713 prev->link = t->link;
2718 barf("unblockThread (I/O): TSO not found");
2722 barf("unblockThread");
2726 tso->link = END_TSO_QUEUE;
2727 tso->why_blocked = NotBlocked;
2728 tso->block_info.closure = NULL;
2729 PUSH_ON_RUN_QUEUE(tso);
2730 RELEASE_LOCK(&sched_mutex);
2734 unblockThread(StgTSO *tso)
2738 ACQUIRE_LOCK(&sched_mutex);
2739 switch (tso->why_blocked) {
2742 return; /* not blocked */
2745 ASSERT(get_itbl(tso->block_info.closure)->type == MVAR);
2747 StgTSO *last_tso = END_TSO_QUEUE;
2748 StgMVar *mvar = (StgMVar *)(tso->block_info.closure);
2751 for (t = mvar->head; t != END_TSO_QUEUE;
2752 last = &t->link, last_tso = t, t = t->link) {
2755 if (mvar->tail == tso) {
2756 mvar->tail = last_tso;
2761 barf("unblockThread (MVAR): TSO not found");
2764 case BlockedOnBlackHole:
2765 ASSERT(get_itbl(tso->block_info.closure)->type == BLACKHOLE_BQ);
2767 StgBlockingQueue *bq = (StgBlockingQueue *)(tso->block_info.closure);
2769 last = &bq->blocking_queue;
2770 for (t = bq->blocking_queue; t != END_TSO_QUEUE;
2771 last = &t->link, t = t->link) {
2777 barf("unblockThread (BLACKHOLE): TSO not found");
2780 case BlockedOnException:
2782 StgTSO *target = tso->block_info.tso;
2784 ASSERT(get_itbl(target)->type == TSO);
2786 while (target->what_next == ThreadRelocated) {
2787 target = target->link;
2788 ASSERT(get_itbl(target)->type == TSO);
2791 ASSERT(target->blocked_exceptions != NULL);
2793 last = &target->blocked_exceptions;
2794 for (t = target->blocked_exceptions; t != END_TSO_QUEUE;
2795 last = &t->link, t = t->link) {
2796 ASSERT(get_itbl(t)->type == TSO);
2802 barf("unblockThread (Exception): TSO not found");
2806 case BlockedOnWrite:
2808 StgTSO *prev = NULL;
2809 for (t = blocked_queue_hd; t != END_TSO_QUEUE;
2810 prev = t, t = t->link) {
2813 blocked_queue_hd = t->link;
2814 if (blocked_queue_tl == t) {
2815 blocked_queue_tl = END_TSO_QUEUE;
2818 prev->link = t->link;
2819 if (blocked_queue_tl == t) {
2820 blocked_queue_tl = prev;
2826 barf("unblockThread (I/O): TSO not found");
2829 case BlockedOnDelay:
2831 StgTSO *prev = NULL;
2832 for (t = sleeping_queue; t != END_TSO_QUEUE;
2833 prev = t, t = t->link) {
2836 sleeping_queue = t->link;
2838 prev->link = t->link;
2843 barf("unblockThread (I/O): TSO not found");
2847 barf("unblockThread");
2851 tso->link = END_TSO_QUEUE;
2852 tso->why_blocked = NotBlocked;
2853 tso->block_info.closure = NULL;
2854 PUSH_ON_RUN_QUEUE(tso);
2855 RELEASE_LOCK(&sched_mutex);
2859 /* -----------------------------------------------------------------------------
2862 * The following function implements the magic for raising an
2863 * asynchronous exception in an existing thread.
2865 * We first remove the thread from any queue on which it might be
2866 * blocked. The possible blockages are MVARs and BLACKHOLE_BQs.
2868 * We strip the stack down to the innermost CATCH_FRAME, building
2869 * thunks in the heap for all the active computations, so they can
2870 * be restarted if necessary. When we reach a CATCH_FRAME, we build
2871 * an application of the handler to the exception, and push it on
2872 * the top of the stack.
2874 * How exactly do we save all the active computations? We create an
2875 * AP_UPD for every UpdateFrame on the stack. Entering one of these
2876 * AP_UPDs pushes everything from the corresponding update frame
2877 * upwards onto the stack. (Actually, it pushes everything up to the
2878 * next update frame plus a pointer to the next AP_UPD object.
2879 * Entering the next AP_UPD object pushes more onto the stack until we
2880 * reach the last AP_UPD object - at which point the stack should look
2881 * exactly as it did when we killed the TSO and we can continue
2882 * execution by entering the closure on top of the stack.
2884 * We can also kill a thread entirely - this happens if either (a) the
2885 * exception passed to raiseAsync is NULL, or (b) there's no
2886 * CATCH_FRAME on the stack. In either case, we strip the entire
2887 * stack and replace the thread with a zombie.
2889 * -------------------------------------------------------------------------- */
2892 deleteThread(StgTSO *tso)
2894 raiseAsync(tso,NULL);
2898 raiseAsync(StgTSO *tso, StgClosure *exception)
2900 StgUpdateFrame* su = tso->su;
2901 StgPtr sp = tso->sp;
2903 /* Thread already dead? */
2904 if (tso->what_next == ThreadComplete || tso->what_next == ThreadKilled) {
2908 IF_DEBUG(scheduler, sched_belch("raising exception in thread %ld.", tso->id));
2910 /* Remove it from any blocking queues */
2913 /* The stack freezing code assumes there's a closure pointer on
2914 * the top of the stack. This isn't always the case with compiled
2915 * code, so we have to push a dummy closure on the top which just
2916 * returns to the next return address on the stack.
2918 if ( LOOKS_LIKE_GHC_INFO((void*)*sp) ) {
2919 *(--sp) = (W_)&stg_dummy_ret_closure;
2923 int words = ((P_)su - (P_)sp) - 1;
2927 /* If we find a CATCH_FRAME, and we've got an exception to raise,
2928 * then build PAP(handler,exception,realworld#), and leave it on
2929 * top of the stack ready to enter.
2931 if (get_itbl(su)->type == CATCH_FRAME && exception != NULL) {
2932 StgCatchFrame *cf = (StgCatchFrame *)su;
2933 /* we've got an exception to raise, so let's pass it to the
2934 * handler in this frame.
2936 ap = (StgAP_UPD *)allocate(sizeofW(StgPAP) + 2);
2937 TICK_ALLOC_UPD_PAP(3,0);
2938 SET_HDR(ap,&stg_PAP_info,cf->header.prof.ccs);
2941 ap->fun = cf->handler; /* :: Exception -> IO a */
2942 ap->payload[0] = exception;
2943 ap->payload[1] = ARG_TAG(0); /* realworld token */
2945 /* throw away the stack from Sp up to and including the
2948 sp = (P_)su + sizeofW(StgCatchFrame) - 1;
2951 /* Restore the blocked/unblocked state for asynchronous exceptions
2952 * at the CATCH_FRAME.
2954 * If exceptions were unblocked at the catch, arrange that they
2955 * are unblocked again after executing the handler by pushing an
2956 * unblockAsyncExceptions_ret stack frame.
2958 if (!cf->exceptions_blocked) {
2959 *(sp--) = (W_)&stg_unblockAsyncExceptionszh_ret_info;
2962 /* Ensure that async exceptions are blocked when running the handler.
2964 if (tso->blocked_exceptions == NULL) {
2965 tso->blocked_exceptions = END_TSO_QUEUE;
2968 /* Put the newly-built PAP on top of the stack, ready to execute
2969 * when the thread restarts.
2973 tso->what_next = ThreadEnterGHC;
2974 IF_DEBUG(sanity, checkTSO(tso));
2978 /* First build an AP_UPD consisting of the stack chunk above the
2979 * current update frame, with the top word on the stack as the
2982 ap = (StgAP_UPD *)allocate(AP_sizeW(words));
2987 ap->fun = (StgClosure *)sp[0];
2989 for(i=0; i < (nat)words; ++i) {
2990 ap->payload[i] = (StgClosure *)*sp++;
2993 switch (get_itbl(su)->type) {
2997 SET_HDR(ap,&stg_AP_UPD_info,su->header.prof.ccs /* ToDo */);
2998 TICK_ALLOC_UP_THK(words+1,0);
3001 fprintf(stderr, "scheduler: Updating ");
3002 printPtr((P_)su->updatee);
3003 fprintf(stderr, " with ");
3004 printObj((StgClosure *)ap);
3007 /* Replace the updatee with an indirection - happily
3008 * this will also wake up any threads currently
3009 * waiting on the result.
3011 * Warning: if we're in a loop, more than one update frame on
3012 * the stack may point to the same object. Be careful not to
3013 * overwrite an IND_OLDGEN in this case, because we'll screw
3014 * up the mutable lists. To be on the safe side, don't
3015 * overwrite any kind of indirection at all. See also
3016 * threadSqueezeStack in GC.c, where we have to make a similar
3019 if (!closure_IND(su->updatee)) {
3020 UPD_IND_NOLOCK(su->updatee,ap); /* revert the black hole */
3023 sp += sizeofW(StgUpdateFrame) -1;
3024 sp[0] = (W_)ap; /* push onto stack */
3030 StgCatchFrame *cf = (StgCatchFrame *)su;
3033 /* We want a PAP, not an AP_UPD. Fortunately, the
3034 * layout's the same.
3036 SET_HDR(ap,&stg_PAP_info,su->header.prof.ccs /* ToDo */);
3037 TICK_ALLOC_UPD_PAP(words+1,0);
3039 /* now build o = FUN(catch,ap,handler) */
3040 o = (StgClosure *)allocate(sizeofW(StgClosure)+2);
3041 TICK_ALLOC_FUN(2,0);
3042 SET_HDR(o,&stg_catch_info,su->header.prof.ccs /* ToDo */);
3043 o->payload[0] = (StgClosure *)ap;
3044 o->payload[1] = cf->handler;
3047 fprintf(stderr, "scheduler: Built ");
3048 printObj((StgClosure *)o);
3051 /* pop the old handler and put o on the stack */
3053 sp += sizeofW(StgCatchFrame) - 1;
3060 StgSeqFrame *sf = (StgSeqFrame *)su;
3063 SET_HDR(ap,&stg_PAP_info,su->header.prof.ccs /* ToDo */);
3064 TICK_ALLOC_UPD_PAP(words+1,0);
3066 /* now build o = FUN(seq,ap) */
3067 o = (StgClosure *)allocate(sizeofW(StgClosure)+1);
3068 TICK_ALLOC_SE_THK(1,0);
3069 SET_HDR(o,&stg_seq_info,su->header.prof.ccs /* ToDo */);
3070 o->payload[0] = (StgClosure *)ap;
3073 fprintf(stderr, "scheduler: Built ");
3074 printObj((StgClosure *)o);
3077 /* pop the old handler and put o on the stack */
3079 sp += sizeofW(StgSeqFrame) - 1;
3085 /* We've stripped the entire stack, the thread is now dead. */
3086 sp += sizeofW(StgStopFrame) - 1;
3087 sp[0] = (W_)exception; /* save the exception */
3088 tso->what_next = ThreadKilled;
3089 tso->su = (StgUpdateFrame *)(sp+1);
3100 /* -----------------------------------------------------------------------------
3101 resurrectThreads is called after garbage collection on the list of
3102 threads found to be garbage. Each of these threads will be woken
3103 up and sent a signal: BlockedOnDeadMVar if the thread was blocked
3104 on an MVar, or NonTermination if the thread was blocked on a Black
3106 -------------------------------------------------------------------------- */
3109 resurrectThreads( StgTSO *threads )
3113 for (tso = threads; tso != END_TSO_QUEUE; tso = next) {
3114 next = tso->global_link;
3115 tso->global_link = all_threads;
3117 IF_DEBUG(scheduler, sched_belch("resurrecting thread %d", tso->id));
3119 switch (tso->why_blocked) {
3121 case BlockedOnException:
3122 raiseAsync(tso,(StgClosure *)BlockedOnDeadMVar_closure);
3124 case BlockedOnBlackHole:
3125 raiseAsync(tso,(StgClosure *)NonTermination_closure);
3128 /* This might happen if the thread was blocked on a black hole
3129 * belonging to a thread that we've just woken up (raiseAsync
3130 * can wake up threads, remember...).
3134 barf("resurrectThreads: thread blocked in a strange way");
3139 /* -----------------------------------------------------------------------------
3140 * Blackhole detection: if we reach a deadlock, test whether any
3141 * threads are blocked on themselves. Any threads which are found to
3142 * be self-blocked get sent a NonTermination exception.
3144 * This is only done in a deadlock situation in order to avoid
3145 * performance overhead in the normal case.
3146 * -------------------------------------------------------------------------- */
3149 detectBlackHoles( void )
3151 StgTSO *t = all_threads;
3152 StgUpdateFrame *frame;
3153 StgClosure *blocked_on;
3155 for (t = all_threads; t != END_TSO_QUEUE; t = t->global_link) {
3157 while (t->what_next == ThreadRelocated) {
3159 ASSERT(get_itbl(t)->type == TSO);
3162 if (t->why_blocked != BlockedOnBlackHole) {
3166 blocked_on = t->block_info.closure;
3168 for (frame = t->su; ; frame = frame->link) {
3169 switch (get_itbl(frame)->type) {
3172 if (frame->updatee == blocked_on) {
3173 /* We are blocking on one of our own computations, so
3174 * send this thread the NonTermination exception.
3177 sched_belch("thread %d is blocked on itself", t->id));
3178 raiseAsync(t, (StgClosure *)NonTermination_closure);
3199 //@node Debugging Routines, Index, Exception Handling Routines, Main scheduling code
3200 //@subsection Debugging Routines
3202 /* -----------------------------------------------------------------------------
3203 Debugging: why is a thread blocked
3204 -------------------------------------------------------------------------- */
3209 printThreadBlockage(StgTSO *tso)
3211 switch (tso->why_blocked) {
3213 fprintf(stderr,"is blocked on read from fd %d", tso->block_info.fd);
3215 case BlockedOnWrite:
3216 fprintf(stderr,"is blocked on write to fd %d", tso->block_info.fd);
3218 case BlockedOnDelay:
3219 fprintf(stderr,"is blocked until %d", tso->block_info.target);
3222 fprintf(stderr,"is blocked on an MVar");
3224 case BlockedOnException:
3225 fprintf(stderr,"is blocked on delivering an exception to thread %d",
3226 tso->block_info.tso->id);
3228 case BlockedOnBlackHole:
3229 fprintf(stderr,"is blocked on a black hole");
3232 fprintf(stderr,"is not blocked");
3236 fprintf(stderr,"is blocked on global address; local FM_BQ is %p (%s)",
3237 tso->block_info.closure, info_type(tso->block_info.closure));
3239 case BlockedOnGA_NoSend:
3240 fprintf(stderr,"is blocked on global address (no send); local FM_BQ is %p (%s)",
3241 tso->block_info.closure, info_type(tso->block_info.closure));
3245 barf("printThreadBlockage: strange tso->why_blocked: %d for TSO %d (%d)",
3246 tso->why_blocked, tso->id, tso);
3251 printThreadStatus(StgTSO *tso)
3253 switch (tso->what_next) {
3255 fprintf(stderr,"has been killed");
3257 case ThreadComplete:
3258 fprintf(stderr,"has completed");
3261 printThreadBlockage(tso);
3266 printAllThreads(void)
3271 char time_string[TIME_STR_LEN], node_str[NODE_STR_LEN];
3272 ullong_format_string(TIME_ON_PROC(CurrentProc),
3273 time_string, rtsFalse/*no commas!*/);
3275 sched_belch("all threads at [%s]:", time_string);
3277 char time_string[TIME_STR_LEN], node_str[NODE_STR_LEN];
3278 ullong_format_string(CURRENT_TIME,
3279 time_string, rtsFalse/*no commas!*/);
3281 sched_belch("all threads at [%s]:", time_string);
3283 sched_belch("all threads:");
3286 for (t = all_threads; t != END_TSO_QUEUE; t = t->global_link) {
3287 fprintf(stderr, "\tthread %d ", t->id);
3288 printThreadStatus(t);
3289 fprintf(stderr,"\n");
3294 Print a whole blocking queue attached to node (debugging only).
3299 print_bq (StgClosure *node)
3301 StgBlockingQueueElement *bqe;
3305 fprintf(stderr,"## BQ of closure %p (%s): ",
3306 node, info_type(node));
3308 /* should cover all closures that may have a blocking queue */
3309 ASSERT(get_itbl(node)->type == BLACKHOLE_BQ ||
3310 get_itbl(node)->type == FETCH_ME_BQ ||
3311 get_itbl(node)->type == RBH ||
3312 get_itbl(node)->type == MVAR);
3314 ASSERT(node!=(StgClosure*)NULL); // sanity check
3316 print_bqe(((StgBlockingQueue*)node)->blocking_queue);
3320 Print a whole blocking queue starting with the element bqe.
3323 print_bqe (StgBlockingQueueElement *bqe)
3328 NB: In a parallel setup a BQ of an RBH must end with an RBH_Save closure;
3330 for (end = (bqe==END_BQ_QUEUE);
3331 !end; // iterate until bqe points to a CONSTR
3332 end = (get_itbl(bqe)->type == CONSTR) || (bqe->link==END_BQ_QUEUE),
3333 bqe = end ? END_BQ_QUEUE : bqe->link) {
3334 ASSERT(bqe != END_BQ_QUEUE); // sanity check
3335 ASSERT(bqe != (StgBlockingQueueElement *)NULL); // sanity check
3336 /* types of closures that may appear in a blocking queue */
3337 ASSERT(get_itbl(bqe)->type == TSO ||
3338 get_itbl(bqe)->type == BLOCKED_FETCH ||
3339 get_itbl(bqe)->type == CONSTR);
3340 /* only BQs of an RBH end with an RBH_Save closure */
3341 //ASSERT(get_itbl(bqe)->type != CONSTR || get_itbl(node)->type == RBH);
3343 switch (get_itbl(bqe)->type) {
3345 fprintf(stderr," TSO %u (%x),",
3346 ((StgTSO *)bqe)->id, ((StgTSO *)bqe));
3349 fprintf(stderr," BF (node=%p, ga=((%x, %d, %x)),",
3350 ((StgBlockedFetch *)bqe)->node,
3351 ((StgBlockedFetch *)bqe)->ga.payload.gc.gtid,
3352 ((StgBlockedFetch *)bqe)->ga.payload.gc.slot,
3353 ((StgBlockedFetch *)bqe)->ga.weight);
3356 fprintf(stderr," %s (IP %p),",
3357 (get_itbl(bqe) == &stg_RBH_Save_0_info ? "RBH_Save_0" :
3358 get_itbl(bqe) == &stg_RBH_Save_1_info ? "RBH_Save_1" :
3359 get_itbl(bqe) == &stg_RBH_Save_2_info ? "RBH_Save_2" :
3360 "RBH_Save_?"), get_itbl(bqe));
3363 barf("Unexpected closure type %s in blocking queue", // of %p (%s)",
3364 info_type((StgClosure *)bqe)); // , node, info_type(node));
3368 fputc('\n', stderr);
3370 # elif defined(GRAN)
3372 print_bq (StgClosure *node)
3374 StgBlockingQueueElement *bqe;
3375 PEs node_loc, tso_loc;
3378 /* should cover all closures that may have a blocking queue */
3379 ASSERT(get_itbl(node)->type == BLACKHOLE_BQ ||
3380 get_itbl(node)->type == FETCH_ME_BQ ||
3381 get_itbl(node)->type == RBH);
3383 ASSERT(node!=(StgClosure*)NULL); // sanity check
3384 node_loc = where_is(node);
3386 fprintf(stderr,"## BQ of closure %p (%s) on [PE %d]: ",
3387 node, info_type(node), node_loc);
3390 NB: In a parallel setup a BQ of an RBH must end with an RBH_Save closure;
3392 for (bqe = ((StgBlockingQueue*)node)->blocking_queue, end = (bqe==END_BQ_QUEUE);
3393 !end; // iterate until bqe points to a CONSTR
3394 end = (get_itbl(bqe)->type == CONSTR) || (bqe->link==END_BQ_QUEUE), bqe = end ? END_BQ_QUEUE : bqe->link) {
3395 ASSERT(bqe != END_BQ_QUEUE); // sanity check
3396 ASSERT(bqe != (StgBlockingQueueElement *)NULL); // sanity check
3397 /* types of closures that may appear in a blocking queue */
3398 ASSERT(get_itbl(bqe)->type == TSO ||
3399 get_itbl(bqe)->type == CONSTR);
3400 /* only BQs of an RBH end with an RBH_Save closure */
3401 ASSERT(get_itbl(bqe)->type != CONSTR || get_itbl(node)->type == RBH);
3403 tso_loc = where_is((StgClosure *)bqe);
3404 switch (get_itbl(bqe)->type) {
3406 fprintf(stderr," TSO %d (%p) on [PE %d],",
3407 ((StgTSO *)bqe)->id, (StgTSO *)bqe, tso_loc);
3410 fprintf(stderr," %s (IP %p),",
3411 (get_itbl(bqe) == &stg_RBH_Save_0_info ? "RBH_Save_0" :
3412 get_itbl(bqe) == &stg_RBH_Save_1_info ? "RBH_Save_1" :
3413 get_itbl(bqe) == &stg_RBH_Save_2_info ? "RBH_Save_2" :
3414 "RBH_Save_?"), get_itbl(bqe));
3417 barf("Unexpected closure type %s in blocking queue of %p (%s)",
3418 info_type((StgClosure *)bqe), node, info_type(node));
3422 fputc('\n', stderr);
3426 Nice and easy: only TSOs on the blocking queue
3429 print_bq (StgClosure *node)
3433 ASSERT(node!=(StgClosure*)NULL); // sanity check
3434 for (tso = ((StgBlockingQueue*)node)->blocking_queue;
3435 tso != END_TSO_QUEUE;
3437 ASSERT(tso!=NULL && tso!=END_TSO_QUEUE); // sanity check
3438 ASSERT(get_itbl(tso)->type == TSO); // guess what, sanity check
3439 fprintf(stderr," TSO %d (%p),", tso->id, tso);
3441 fputc('\n', stderr);
3452 for (i=0, tso=run_queue_hd;
3453 tso != END_TSO_QUEUE;
3462 sched_belch(char *s, ...)
3467 fprintf(stderr, "scheduler (task %ld): ", pthread_self());
3469 fprintf(stderr, "== ");
3471 fprintf(stderr, "scheduler: ");
3473 vfprintf(stderr, s, ap);
3474 fprintf(stderr, "\n");
3480 //@node Index, , Debugging Routines, Main scheduling code
3484 //* MainRegTable:: @cindex\s-+MainRegTable
3485 //* StgMainThread:: @cindex\s-+StgMainThread
3486 //* awaken_blocked_queue:: @cindex\s-+awaken_blocked_queue
3487 //* blocked_queue_hd:: @cindex\s-+blocked_queue_hd
3488 //* blocked_queue_tl:: @cindex\s-+blocked_queue_tl
3489 //* context_switch:: @cindex\s-+context_switch
3490 //* createThread:: @cindex\s-+createThread
3491 //* free_capabilities:: @cindex\s-+free_capabilities
3492 //* gc_pending_cond:: @cindex\s-+gc_pending_cond
3493 //* initScheduler:: @cindex\s-+initScheduler
3494 //* interrupted:: @cindex\s-+interrupted
3495 //* n_free_capabilities:: @cindex\s-+n_free_capabilities
3496 //* next_thread_id:: @cindex\s-+next_thread_id
3497 //* print_bq:: @cindex\s-+print_bq
3498 //* run_queue_hd:: @cindex\s-+run_queue_hd
3499 //* run_queue_tl:: @cindex\s-+run_queue_tl
3500 //* sched_mutex:: @cindex\s-+sched_mutex
3501 //* schedule:: @cindex\s-+schedule
3502 //* take_off_run_queue:: @cindex\s-+take_off_run_queue
3503 //* task_ids:: @cindex\s-+task_ids
3504 //* term_mutex:: @cindex\s-+term_mutex
3505 //* thread_ready_cond:: @cindex\s-+thread_ready_cond