10 #include "snapshot-interface.h"
12 #include "clockvector.h"
13 #include "cyclegraph.h"
16 #include "threads-model.h"
19 #define INITIAL_THREAD_ID 0
24 bug_message(const char *str) {
25 const char *fmt = " [BUG] %s\n";
26 msg = (char *)snapshot_malloc(strlen(fmt) + strlen(str));
27 sprintf(msg, fmt, str);
29 ~bug_message() { if (msg) snapshot_free(msg); }
32 void print() { model_print("%s", msg); }
38 * Structure for holding small ModelChecker members that should be snapshotted
40 struct model_snapshot_members {
41 model_snapshot_members() :
43 /* First thread created will have id INITIAL_THREAD_ID */
44 next_thread_id(INITIAL_THREAD_ID),
45 used_sequence_numbers(0),
49 failed_promise(false),
50 too_many_reads(false),
51 bad_synchronization(false),
55 ~model_snapshot_members() {
56 for (unsigned int i = 0; i < bugs.size(); i++)
61 ModelAction *current_action;
62 unsigned int next_thread_id;
63 modelclock_t used_sequence_numbers;
64 ModelAction *next_backtrack;
65 std::vector< bug_message *, SnapshotAlloc<bug_message *> > bugs;
66 struct execution_stats stats;
69 /** @brief Incorrectly-ordered synchronization was made */
70 bool bad_synchronization;
76 /** @brief Constructor */
77 ModelChecker::ModelChecker(struct model_params params) :
78 /* Initialize default scheduler */
80 scheduler(new Scheduler()),
82 earliest_diverge(NULL),
83 action_trace(new action_list_t()),
84 thread_map(new HashTable<int, Thread *, int>()),
85 obj_map(new HashTable<const void *, action_list_t *, uintptr_t, 4>()),
86 lock_waiters_map(new HashTable<const void *, action_list_t *, uintptr_t, 4>()),
87 condvar_waiters_map(new HashTable<const void *, action_list_t *, uintptr_t, 4>()),
88 obj_thrd_map(new HashTable<void *, std::vector<action_list_t> *, uintptr_t, 4 >()),
89 promises(new std::vector< Promise *, SnapshotAlloc<Promise *> >()),
90 futurevalues(new std::vector< struct PendingFutureValue, SnapshotAlloc<struct PendingFutureValue> >()),
91 pending_rel_seqs(new std::vector< struct release_seq *, SnapshotAlloc<struct release_seq *> >()),
92 thrd_last_action(new std::vector< ModelAction *, SnapshotAlloc<ModelAction *> >(1)),
93 thrd_last_fence_release(new std::vector< ModelAction *, SnapshotAlloc<ModelAction *> >()),
94 node_stack(new NodeStack()),
95 priv(new struct model_snapshot_members()),
96 mo_graph(new CycleGraph())
98 /* Initialize a model-checker thread, for special ModelActions */
99 model_thread = new Thread(get_next_id());
100 thread_map->put(id_to_int(model_thread->get_id()), model_thread);
103 /** @brief Destructor */
104 ModelChecker::~ModelChecker()
106 for (unsigned int i = 0; i < get_num_threads(); i++)
107 delete thread_map->get(i);
112 delete lock_waiters_map;
113 delete condvar_waiters_map;
116 for (unsigned int i = 0; i < promises->size(); i++)
117 delete (*promises)[i];
120 delete pending_rel_seqs;
122 delete thrd_last_action;
123 delete thrd_last_fence_release;
130 static action_list_t * get_safe_ptr_action(HashTable<const void *, action_list_t *, uintptr_t, 4> * hash, void * ptr)
132 action_list_t *tmp = hash->get(ptr);
134 tmp = new action_list_t();
140 static std::vector<action_list_t> * get_safe_ptr_vect_action(HashTable<void *, std::vector<action_list_t> *, uintptr_t, 4> * hash, void * ptr)
142 std::vector<action_list_t> *tmp = hash->get(ptr);
144 tmp = new std::vector<action_list_t>();
151 * Restores user program to initial state and resets all model-checker data
154 void ModelChecker::reset_to_initial_state()
156 DEBUG("+++ Resetting to initial state +++\n");
157 node_stack->reset_execution();
159 /* Print all model-checker output before rollback */
162 snapshot_backtrack_before(0);
165 /** @return a thread ID for a new Thread */
166 thread_id_t ModelChecker::get_next_id()
168 return priv->next_thread_id++;
171 /** @return the number of user threads created during this execution */
172 unsigned int ModelChecker::get_num_threads() const
174 return priv->next_thread_id;
178 * Must be called from user-thread context (e.g., through the global
179 * thread_current() interface)
181 * @return The currently executing Thread.
183 Thread * ModelChecker::get_current_thread() const
185 return scheduler->get_current_thread();
188 /** @return a sequence number for a new ModelAction */
189 modelclock_t ModelChecker::get_next_seq_num()
191 return ++priv->used_sequence_numbers;
194 Node * ModelChecker::get_curr_node() const
196 return node_stack->get_head();
200 * @brief Choose the next thread to execute.
202 * This function chooses the next thread that should execute. It can force the
203 * adjacency of read/write portions of a RMW action, force THREAD_CREATE to be
204 * followed by a THREAD_START, or it can enforce execution replay/backtracking.
205 * The model-checker may have no preference regarding the next thread (i.e.,
206 * when exploring a new execution ordering), in which case this will return
208 * @param curr The current ModelAction. This action might guide the choice of
210 * @return The next thread to run. If the model-checker has no preference, NULL.
212 Thread * ModelChecker::get_next_thread(ModelAction *curr)
217 /* Do not split atomic actions. */
219 return thread_current();
220 else if (curr->get_type() == THREAD_CREATE)
221 return curr->get_thread_operand();
224 /* Have we completed exploring the preselected path? */
228 /* Else, we are trying to replay an execution */
229 ModelAction *next = node_stack->get_next()->get_action();
231 if (next == diverge) {
232 if (earliest_diverge == NULL || *diverge < *earliest_diverge)
233 earliest_diverge = diverge;
235 Node *nextnode = next->get_node();
236 Node *prevnode = nextnode->get_parent();
237 scheduler->update_sleep_set(prevnode);
239 /* Reached divergence point */
240 if (nextnode->increment_misc()) {
241 /* The next node will try to satisfy a different misc_index values. */
242 tid = next->get_tid();
243 node_stack->pop_restofstack(2);
244 } else if (nextnode->increment_promise()) {
245 /* The next node will try to satisfy a different set of promises. */
246 tid = next->get_tid();
247 node_stack->pop_restofstack(2);
248 } else if (nextnode->increment_read_from()) {
249 /* The next node will read from a different value. */
250 tid = next->get_tid();
251 node_stack->pop_restofstack(2);
252 } else if (nextnode->increment_future_value()) {
253 /* The next node will try to read from a different future value. */
254 tid = next->get_tid();
255 node_stack->pop_restofstack(2);
256 } else if (nextnode->increment_relseq_break()) {
257 /* The next node will try to resolve a release sequence differently */
258 tid = next->get_tid();
259 node_stack->pop_restofstack(2);
262 /* Make a different thread execute for next step */
263 scheduler->add_sleep(get_thread(next->get_tid()));
264 tid = prevnode->get_next_backtrack();
265 /* Make sure the backtracked thread isn't sleeping. */
266 node_stack->pop_restofstack(1);
267 if (diverge == earliest_diverge) {
268 earliest_diverge = prevnode->get_action();
271 /* The correct sleep set is in the parent node. */
274 DEBUG("*** Divergence point ***\n");
278 tid = next->get_tid();
280 DEBUG("*** ModelChecker chose next thread = %d ***\n", id_to_int(tid));
281 ASSERT(tid != THREAD_ID_T_NONE);
282 return thread_map->get(id_to_int(tid));
286 * We need to know what the next actions of all threads in the sleep
287 * set will be. This method computes them and stores the actions at
288 * the corresponding thread object's pending action.
291 void ModelChecker::execute_sleep_set()
293 for (unsigned int i = 0; i < get_num_threads(); i++) {
294 thread_id_t tid = int_to_id(i);
295 Thread *thr = get_thread(tid);
296 if (scheduler->is_sleep_set(thr) && thr->get_pending() == NULL) {
297 thr->set_state(THREAD_RUNNING);
298 scheduler->next_thread(thr);
299 Thread::swap(&system_context, thr);
300 priv->current_action->set_sleep_flag();
301 thr->set_pending(priv->current_action);
306 void ModelChecker::wake_up_sleeping_actions(ModelAction *curr)
308 for (unsigned int i = 0; i < get_num_threads(); i++) {
309 Thread *thr = get_thread(int_to_id(i));
310 if (scheduler->is_sleep_set(thr)) {
311 ModelAction *pending_act = thr->get_pending();
312 if ((!curr->is_rmwr()) && pending_act->could_synchronize_with(curr))
313 //Remove this thread from sleep set
314 scheduler->remove_sleep(thr);
319 /** @brief Alert the model-checker that an incorrectly-ordered
320 * synchronization was made */
321 void ModelChecker::set_bad_synchronization()
323 priv->bad_synchronization = true;
326 bool ModelChecker::has_asserted() const
328 return priv->asserted;
331 void ModelChecker::set_assert()
333 priv->asserted = true;
337 * Check if we are in a deadlock. Should only be called at the end of an
338 * execution, although it should not give false positives in the middle of an
339 * execution (there should be some ENABLED thread).
341 * @return True if program is in a deadlock; false otherwise
343 bool ModelChecker::is_deadlocked() const
345 bool blocking_threads = false;
346 for (unsigned int i = 0; i < get_num_threads(); i++) {
347 thread_id_t tid = int_to_id(i);
350 Thread *t = get_thread(tid);
351 if (!t->is_model_thread() && t->get_pending())
352 blocking_threads = true;
354 return blocking_threads;
358 * Check if this is a complete execution. That is, have all thread completed
359 * execution (rather than exiting because sleep sets have forced a redundant
362 * @return True if the execution is complete.
364 bool ModelChecker::is_complete_execution() const
366 for (unsigned int i = 0; i < get_num_threads(); i++)
367 if (is_enabled(int_to_id(i)))
373 * @brief Assert a bug in the executing program.
375 * Use this function to assert any sort of bug in the user program. If the
376 * current trace is feasible (actually, a prefix of some feasible execution),
377 * then this execution will be aborted, printing the appropriate message. If
378 * the current trace is not yet feasible, the error message will be stashed and
379 * printed if the execution ever becomes feasible.
381 * @param msg Descriptive message for the bug (do not include newline char)
382 * @return True if bug is immediately-feasible
384 bool ModelChecker::assert_bug(const char *msg)
386 priv->bugs.push_back(new bug_message(msg));
388 if (isfeasibleprefix()) {
396 * @brief Assert a bug in the executing program, asserted by a user thread
397 * @see ModelChecker::assert_bug
398 * @param msg Descriptive message for the bug (do not include newline char)
400 void ModelChecker::assert_user_bug(const char *msg)
402 /* If feasible bug, bail out now */
404 switch_to_master(NULL);
407 /** @return True, if any bugs have been reported for this execution */
408 bool ModelChecker::have_bug_reports() const
410 return priv->bugs.size() != 0;
413 /** @brief Print bug report listing for this execution (if any bugs exist) */
414 void ModelChecker::print_bugs() const
416 if (have_bug_reports()) {
417 model_print("Bug report: %zu bug%s detected\n",
419 priv->bugs.size() > 1 ? "s" : "");
420 for (unsigned int i = 0; i < priv->bugs.size(); i++)
421 priv->bugs[i]->print();
426 * @brief Record end-of-execution stats
428 * Must be run when exiting an execution. Records various stats.
429 * @see struct execution_stats
431 void ModelChecker::record_stats()
434 if (!isfeasibleprefix())
435 stats.num_infeasible++;
436 else if (have_bug_reports())
437 stats.num_buggy_executions++;
438 else if (is_complete_execution())
439 stats.num_complete++;
441 stats.num_redundant++;
444 /** @brief Print execution stats */
445 void ModelChecker::print_stats() const
447 model_print("Number of complete, bug-free executions: %d\n", stats.num_complete);
448 model_print("Number of redundant executions: %d\n", stats.num_redundant);
449 model_print("Number of buggy executions: %d\n", stats.num_buggy_executions);
450 model_print("Number of infeasible executions: %d\n", stats.num_infeasible);
451 model_print("Total executions: %d\n", stats.num_total);
452 model_print("Total nodes created: %d\n", node_stack->get_total_nodes());
456 * @brief End-of-exeuction print
457 * @param printbugs Should any existing bugs be printed?
459 void ModelChecker::print_execution(bool printbugs) const
461 print_program_output();
463 if (DBG_ENABLED() || params.verbose) {
464 model_print("Earliest divergence point since last feasible execution:\n");
465 if (earliest_diverge)
466 earliest_diverge->print();
468 model_print("(Not set)\n");
474 /* Don't print invalid bugs */
483 * Queries the model-checker for more executions to explore and, if one
484 * exists, resets the model-checker state to execute a new execution.
486 * @return If there are more executions to explore, return true. Otherwise,
489 bool ModelChecker::next_execution()
492 /* Is this execution a feasible execution that's worth bug-checking? */
493 bool complete = isfeasibleprefix() && (is_complete_execution() ||
496 /* End-of-execution bug checks */
499 assert_bug("Deadlock detected");
507 if (DBG_ENABLED() || params.verbose || (complete && have_bug_reports()))
508 print_execution(complete);
510 clear_program_output();
513 earliest_diverge = NULL;
515 if ((diverge = get_next_backtrack()) == NULL)
519 model_print("Next execution will diverge at:\n");
523 reset_to_initial_state();
527 ModelAction * ModelChecker::get_last_conflict(ModelAction *act)
529 switch (act->get_type()) {
534 /* Optimization: relaxed operations don't need backtracking */
535 if (act->is_relaxed())
537 /* linear search: from most recent to oldest */
538 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
539 action_list_t::reverse_iterator rit;
540 for (rit = list->rbegin(); rit != list->rend(); rit++) {
541 ModelAction *prev = *rit;
542 if (prev->could_synchronize_with(act))
548 case ATOMIC_TRYLOCK: {
549 /* linear search: from most recent to oldest */
550 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
551 action_list_t::reverse_iterator rit;
552 for (rit = list->rbegin(); rit != list->rend(); rit++) {
553 ModelAction *prev = *rit;
554 if (act->is_conflicting_lock(prev))
559 case ATOMIC_UNLOCK: {
560 /* linear search: from most recent to oldest */
561 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
562 action_list_t::reverse_iterator rit;
563 for (rit = list->rbegin(); rit != list->rend(); rit++) {
564 ModelAction *prev = *rit;
565 if (!act->same_thread(prev) && prev->is_failed_trylock())
571 /* linear search: from most recent to oldest */
572 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
573 action_list_t::reverse_iterator rit;
574 for (rit = list->rbegin(); rit != list->rend(); rit++) {
575 ModelAction *prev = *rit;
576 if (!act->same_thread(prev) && prev->is_failed_trylock())
578 if (!act->same_thread(prev) && prev->is_notify())
584 case ATOMIC_NOTIFY_ALL:
585 case ATOMIC_NOTIFY_ONE: {
586 /* linear search: from most recent to oldest */
587 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
588 action_list_t::reverse_iterator rit;
589 for (rit = list->rbegin(); rit != list->rend(); rit++) {
590 ModelAction *prev = *rit;
591 if (!act->same_thread(prev) && prev->is_wait())
602 /** This method finds backtracking points where we should try to
603 * reorder the parameter ModelAction against.
605 * @param the ModelAction to find backtracking points for.
607 void ModelChecker::set_backtracking(ModelAction *act)
609 Thread *t = get_thread(act);
610 ModelAction *prev = get_last_conflict(act);
614 Node *node = prev->get_node()->get_parent();
616 int low_tid, high_tid;
617 if (node->enabled_status(t->get_id()) == THREAD_ENABLED) {
618 low_tid = id_to_int(act->get_tid());
619 high_tid = low_tid + 1;
622 high_tid = get_num_threads();
625 for (int i = low_tid; i < high_tid; i++) {
626 thread_id_t tid = int_to_id(i);
628 /* Make sure this thread can be enabled here. */
629 if (i >= node->get_num_threads())
632 /* Don't backtrack into a point where the thread is disabled or sleeping. */
633 if (node->enabled_status(tid) != THREAD_ENABLED)
636 /* Check if this has been explored already */
637 if (node->has_been_explored(tid))
640 /* See if fairness allows */
641 if (model->params.fairwindow != 0 && !node->has_priority(tid)) {
643 for (int t = 0; t < node->get_num_threads(); t++) {
644 thread_id_t tother = int_to_id(t);
645 if (node->is_enabled(tother) && node->has_priority(tother)) {
653 /* Cache the latest backtracking point */
654 set_latest_backtrack(prev);
656 /* If this is a new backtracking point, mark the tree */
657 if (!node->set_backtrack(tid))
659 DEBUG("Setting backtrack: conflict = %d, instead tid = %d\n",
660 id_to_int(prev->get_tid()),
661 id_to_int(t->get_id()));
670 * @brief Cache the a backtracking point as the "most recent", if eligible
672 * Note that this does not prepare the NodeStack for this backtracking
673 * operation, it only caches the action on a per-execution basis
675 * @param act The operation at which we should explore a different next action
676 * (i.e., backtracking point)
677 * @return True, if this action is now the most recent backtracking point;
680 bool ModelChecker::set_latest_backtrack(ModelAction *act)
682 if (!priv->next_backtrack || *act > *priv->next_backtrack) {
683 priv->next_backtrack = act;
690 * Returns last backtracking point. The model checker will explore a different
691 * path for this point in the next execution.
692 * @return The ModelAction at which the next execution should diverge.
694 ModelAction * ModelChecker::get_next_backtrack()
696 ModelAction *next = priv->next_backtrack;
697 priv->next_backtrack = NULL;
702 * Processes a read or rmw model action.
703 * @param curr is the read model action to process.
704 * @param second_part_of_rmw is boolean that is true is this is the second action of a rmw.
705 * @return True if processing this read updates the mo_graph.
707 bool ModelChecker::process_read(ModelAction *curr, bool second_part_of_rmw)
709 uint64_t value = VALUE_NONE;
710 bool updated = false;
712 const ModelAction *reads_from = curr->get_node()->get_read_from();
713 if (reads_from != NULL) {
714 mo_graph->startChanges();
716 value = reads_from->get_value();
717 bool r_status = false;
719 if (!second_part_of_rmw) {
720 check_recency(curr, reads_from);
721 r_status = r_modification_order(curr, reads_from);
725 if (!second_part_of_rmw && is_infeasible() && (curr->get_node()->increment_read_from() || curr->get_node()->increment_future_value())) {
726 mo_graph->rollbackChanges();
727 priv->too_many_reads = false;
731 read_from(curr, reads_from);
732 mo_graph->commitChanges();
733 mo_check_promises(curr->get_tid(), reads_from, NULL);
736 } else if (!second_part_of_rmw) {
737 /* Read from future value */
738 struct future_value fv = curr->get_node()->get_future_value();
740 curr->set_value(fv.value);
741 curr->set_read_from(NULL);
742 promises->push_back(new Promise(curr, fv));
744 get_thread(curr)->set_return_value(value);
750 * Processes a lock, trylock, or unlock model action. @param curr is
751 * the read model action to process.
753 * The try lock operation checks whether the lock is taken. If not,
754 * it falls to the normal lock operation case. If so, it returns
757 * The lock operation has already been checked that it is enabled, so
758 * it just grabs the lock and synchronizes with the previous unlock.
760 * The unlock operation has to re-enable all of the threads that are
761 * waiting on the lock.
763 * @return True if synchronization was updated; false otherwise
765 bool ModelChecker::process_mutex(ModelAction *curr)
767 std::mutex *mutex = NULL;
768 struct std::mutex_state *state = NULL;
770 if (curr->is_trylock() || curr->is_lock() || curr->is_unlock()) {
771 mutex = (std::mutex *)curr->get_location();
772 state = mutex->get_state();
773 } else if (curr->is_wait()) {
774 mutex = (std::mutex *)curr->get_value();
775 state = mutex->get_state();
778 switch (curr->get_type()) {
779 case ATOMIC_TRYLOCK: {
780 bool success = !state->islocked;
781 curr->set_try_lock(success);
783 get_thread(curr)->set_return_value(0);
786 get_thread(curr)->set_return_value(1);
788 //otherwise fall into the lock case
790 if (curr->get_cv()->getClock(state->alloc_tid) <= state->alloc_clock)
791 assert_bug("Lock access before initialization");
792 state->islocked = true;
793 ModelAction *unlock = get_last_unlock(curr);
794 //synchronize with the previous unlock statement
795 if (unlock != NULL) {
796 curr->synchronize_with(unlock);
801 case ATOMIC_UNLOCK: {
803 state->islocked = false;
804 //wake up the other threads
805 action_list_t *waiters = get_safe_ptr_action(lock_waiters_map, curr->get_location());
806 //activate all the waiting threads
807 for (action_list_t::iterator rit = waiters->begin(); rit != waiters->end(); rit++) {
808 scheduler->wake(get_thread(*rit));
815 state->islocked = false;
816 //wake up the other threads
817 action_list_t *waiters = get_safe_ptr_action(lock_waiters_map, (void *) curr->get_value());
818 //activate all the waiting threads
819 for (action_list_t::iterator rit = waiters->begin(); rit != waiters->end(); rit++) {
820 scheduler->wake(get_thread(*rit));
823 //check whether we should go to sleep or not...simulate spurious failures
824 if (curr->get_node()->get_misc() == 0) {
825 get_safe_ptr_action(condvar_waiters_map, curr->get_location())->push_back(curr);
827 scheduler->sleep(get_thread(curr));
831 case ATOMIC_NOTIFY_ALL: {
832 action_list_t *waiters = get_safe_ptr_action(condvar_waiters_map, curr->get_location());
833 //activate all the waiting threads
834 for (action_list_t::iterator rit = waiters->begin(); rit != waiters->end(); rit++) {
835 scheduler->wake(get_thread(*rit));
840 case ATOMIC_NOTIFY_ONE: {
841 action_list_t *waiters = get_safe_ptr_action(condvar_waiters_map, curr->get_location());
842 int wakeupthread = curr->get_node()->get_misc();
843 action_list_t::iterator it = waiters->begin();
844 advance(it, wakeupthread);
845 scheduler->wake(get_thread(*it));
856 void ModelChecker::add_future_value(const ModelAction *writer, ModelAction *reader)
858 /* Do more ambitious checks now that mo is more complete */
859 if (mo_may_allow(writer, reader)) {
860 Node *node = reader->get_node();
862 /* Find an ancestor thread which exists at the time of the reader */
863 Thread *write_thread = get_thread(writer);
864 while (id_to_int(write_thread->get_id()) >= node->get_num_threads())
865 write_thread = write_thread->get_parent();
867 struct future_value fv = {
869 writer->get_seq_number() + params.maxfuturedelay,
870 write_thread->get_id(),
872 if (node->add_future_value(fv))
873 set_latest_backtrack(reader);
878 * Process a write ModelAction
879 * @param curr The ModelAction to process
880 * @return True if the mo_graph was updated or promises were resolved
882 bool ModelChecker::process_write(ModelAction *curr)
884 bool updated_mod_order = w_modification_order(curr);
885 bool updated_promises = resolve_promises(curr);
887 if (promises->size() == 0) {
888 for (unsigned int i = 0; i < futurevalues->size(); i++) {
889 struct PendingFutureValue pfv = (*futurevalues)[i];
890 add_future_value(pfv.writer, pfv.act);
892 futurevalues->clear();
895 mo_graph->commitChanges();
896 mo_check_promises(curr->get_tid(), curr, NULL);
898 get_thread(curr)->set_return_value(VALUE_NONE);
899 return updated_mod_order || updated_promises;
903 * Process a fence ModelAction
904 * @param curr The ModelAction to process
905 * @return True if synchronization was updated
907 bool ModelChecker::process_fence(ModelAction *curr)
910 * fence-relaxed: no-op
911 * fence-release: only log the occurence (not in this function), for
912 * use in later synchronization
913 * fence-acquire (this function): search for hypothetical release
916 bool updated = false;
917 if (curr->is_acquire()) {
918 action_list_t *list = action_trace;
919 action_list_t::reverse_iterator rit;
920 /* Find X : is_read(X) && X --sb-> curr */
921 for (rit = list->rbegin(); rit != list->rend(); rit++) {
922 ModelAction *act = *rit;
925 if (act->get_tid() != curr->get_tid())
927 /* Stop at the beginning of the thread */
928 if (act->is_thread_start())
930 /* Stop once we reach a prior fence-acquire */
931 if (act->is_fence() && act->is_acquire())
935 /* read-acquire will find its own release sequences */
936 if (act->is_acquire())
939 /* Establish hypothetical release sequences */
940 rel_heads_list_t release_heads;
941 get_release_seq_heads(curr, act, &release_heads);
942 for (unsigned int i = 0; i < release_heads.size(); i++)
943 if (!curr->synchronize_with(release_heads[i]))
944 set_bad_synchronization();
945 if (release_heads.size() != 0)
953 * @brief Process the current action for thread-related activity
955 * Performs current-action processing for a THREAD_* ModelAction. Proccesses
956 * may include setting Thread status, completing THREAD_FINISH/THREAD_JOIN
957 * synchronization, etc. This function is a no-op for non-THREAD actions
958 * (e.g., ATOMIC_{READ,WRITE,RMW,LOCK}, etc.)
960 * @param curr The current action
961 * @return True if synchronization was updated or a thread completed
963 bool ModelChecker::process_thread_action(ModelAction *curr)
965 bool updated = false;
967 switch (curr->get_type()) {
968 case THREAD_CREATE: {
969 Thread *th = curr->get_thread_operand();
970 th->set_creation(curr);
971 /* Promises can be satisfied by children */
972 for (unsigned int i = 0; i < promises->size(); i++) {
973 Promise *promise = (*promises)[i];
974 if (promise->thread_is_available(curr->get_tid()))
975 promise->add_thread(th->get_id());
980 Thread *blocking = curr->get_thread_operand();
981 ModelAction *act = get_last_action(blocking->get_id());
982 curr->synchronize_with(act);
983 updated = true; /* trigger rel-seq checks */
986 case THREAD_FINISH: {
987 Thread *th = get_thread(curr);
988 while (!th->wait_list_empty()) {
989 ModelAction *act = th->pop_wait_list();
990 scheduler->wake(get_thread(act));
993 /* Completed thread can't satisfy promises */
994 for (unsigned int i = 0; i < promises->size(); i++) {
995 Promise *promise = (*promises)[i];
996 if (promise->thread_is_available(th->get_id()))
997 if (promise->eliminate_thread(th->get_id()))
998 priv->failed_promise = true;
1000 updated = true; /* trigger rel-seq checks */
1003 case THREAD_START: {
1004 check_promises(curr->get_tid(), NULL, curr->get_cv());
1015 * @brief Process the current action for release sequence fixup activity
1017 * Performs model-checker release sequence fixups for the current action,
1018 * forcing a single pending release sequence to break (with a given, potential
1019 * "loose" write) or to complete (i.e., synchronize). If a pending release
1020 * sequence forms a complete release sequence, then we must perform the fixup
1021 * synchronization, mo_graph additions, etc.
1023 * @param curr The current action; must be a release sequence fixup action
1024 * @param work_queue The work queue to which to add work items as they are
1027 void ModelChecker::process_relseq_fixup(ModelAction *curr, work_queue_t *work_queue)
1029 const ModelAction *write = curr->get_node()->get_relseq_break();
1030 struct release_seq *sequence = pending_rel_seqs->back();
1031 pending_rel_seqs->pop_back();
1033 ModelAction *acquire = sequence->acquire;
1034 const ModelAction *rf = sequence->rf;
1035 const ModelAction *release = sequence->release;
1039 ASSERT(release->same_thread(rf));
1041 if (write == NULL) {
1043 * @todo Forcing a synchronization requires that we set
1044 * modification order constraints. For instance, we can't allow
1045 * a fixup sequence in which two separate read-acquire
1046 * operations read from the same sequence, where the first one
1047 * synchronizes and the other doesn't. Essentially, we can't
1048 * allow any writes to insert themselves between 'release' and
1052 /* Must synchronize */
1053 if (!acquire->synchronize_with(release)) {
1054 set_bad_synchronization();
1057 /* Re-check all pending release sequences */
1058 work_queue->push_back(CheckRelSeqWorkEntry(NULL));
1059 /* Re-check act for mo_graph edges */
1060 work_queue->push_back(MOEdgeWorkEntry(acquire));
1062 /* propagate synchronization to later actions */
1063 action_list_t::reverse_iterator rit = action_trace->rbegin();
1064 for (; (*rit) != acquire; rit++) {
1065 ModelAction *propagate = *rit;
1066 if (acquire->happens_before(propagate)) {
1067 propagate->synchronize_with(acquire);
1068 /* Re-check 'propagate' for mo_graph edges */
1069 work_queue->push_back(MOEdgeWorkEntry(propagate));
1073 /* Break release sequence with new edges:
1074 * release --mo--> write --mo--> rf */
1075 mo_graph->addEdge(release, write);
1076 mo_graph->addEdge(write, rf);
1079 /* See if we have realized a data race */
1084 * Initialize the current action by performing one or more of the following
1085 * actions, as appropriate: merging RMWR and RMWC/RMW actions, stepping forward
1086 * in the NodeStack, manipulating backtracking sets, allocating and
1087 * initializing clock vectors, and computing the promises to fulfill.
1089 * @param curr The current action, as passed from the user context; may be
1090 * freed/invalidated after the execution of this function, with a different
1091 * action "returned" its place (pass-by-reference)
1092 * @return True if curr is a newly-explored action; false otherwise
1094 bool ModelChecker::initialize_curr_action(ModelAction **curr)
1096 ModelAction *newcurr;
1098 if ((*curr)->is_rmwc() || (*curr)->is_rmw()) {
1099 newcurr = process_rmw(*curr);
1102 if (newcurr->is_rmw())
1103 compute_promises(newcurr);
1109 (*curr)->set_seq_number(get_next_seq_num());
1111 newcurr = node_stack->explore_action(*curr, scheduler->get_enabled_array());
1113 /* First restore type and order in case of RMW operation */
1114 if ((*curr)->is_rmwr())
1115 newcurr->copy_typeandorder(*curr);
1117 ASSERT((*curr)->get_location() == newcurr->get_location());
1118 newcurr->copy_from_new(*curr);
1120 /* Discard duplicate ModelAction; use action from NodeStack */
1123 /* Always compute new clock vector */
1124 newcurr->create_cv(get_parent_action(newcurr->get_tid()));
1127 return false; /* Action was explored previously */
1131 /* Always compute new clock vector */
1132 newcurr->create_cv(get_parent_action(newcurr->get_tid()));
1134 /* Assign most recent release fence */
1135 newcurr->set_last_fence_release(get_last_fence_release(newcurr->get_tid()));
1138 * Perform one-time actions when pushing new ModelAction onto
1141 if (newcurr->is_write())
1142 compute_promises(newcurr);
1143 else if (newcurr->is_relseq_fixup())
1144 compute_relseq_breakwrites(newcurr);
1145 else if (newcurr->is_wait())
1146 newcurr->get_node()->set_misc_max(2);
1147 else if (newcurr->is_notify_one()) {
1148 newcurr->get_node()->set_misc_max(get_safe_ptr_action(condvar_waiters_map, newcurr->get_location())->size());
1150 return true; /* This was a new ModelAction */
1155 * @brief Establish reads-from relation between two actions
1157 * Perform basic operations involved with establishing a concrete rf relation,
1158 * including setting the ModelAction data and checking for release sequences.
1160 * @param act The action that is reading (must be a read)
1161 * @param rf The action from which we are reading (must be a write)
1163 * @return True if this read established synchronization
1165 bool ModelChecker::read_from(ModelAction *act, const ModelAction *rf)
1167 act->set_read_from(rf);
1168 if (rf != NULL && act->is_acquire()) {
1169 rel_heads_list_t release_heads;
1170 get_release_seq_heads(act, act, &release_heads);
1171 int num_heads = release_heads.size();
1172 for (unsigned int i = 0; i < release_heads.size(); i++)
1173 if (!act->synchronize_with(release_heads[i])) {
1174 set_bad_synchronization();
1177 return num_heads > 0;
1183 * @brief Check whether a model action is enabled.
1185 * Checks whether a lock or join operation would be successful (i.e., is the
1186 * lock already locked, or is the joined thread already complete). If not, put
1187 * the action in a waiter list.
1189 * @param curr is the ModelAction to check whether it is enabled.
1190 * @return a bool that indicates whether the action is enabled.
1192 bool ModelChecker::check_action_enabled(ModelAction *curr) {
1193 if (curr->is_lock()) {
1194 std::mutex *lock = (std::mutex *)curr->get_location();
1195 struct std::mutex_state *state = lock->get_state();
1196 if (state->islocked) {
1197 //Stick the action in the appropriate waiting queue
1198 get_safe_ptr_action(lock_waiters_map, curr->get_location())->push_back(curr);
1201 } else if (curr->get_type() == THREAD_JOIN) {
1202 Thread *blocking = (Thread *)curr->get_location();
1203 if (!blocking->is_complete()) {
1204 blocking->push_wait_list(curr);
1213 * Stores the ModelAction for the current thread action. Call this
1214 * immediately before switching from user- to system-context to pass
1215 * data between them.
1216 * @param act The ModelAction created by the user-thread action
1218 void ModelChecker::set_current_action(ModelAction *act) {
1219 priv->current_action = act;
1223 * This is the heart of the model checker routine. It performs model-checking
1224 * actions corresponding to a given "current action." Among other processes, it
1225 * calculates reads-from relationships, updates synchronization clock vectors,
1226 * forms a memory_order constraints graph, and handles replay/backtrack
1227 * execution when running permutations of previously-observed executions.
1229 * @param curr The current action to process
1230 * @return The ModelAction that is actually executed; may be different than
1231 * curr; may be NULL, if the current action is not enabled to run
1233 ModelAction * ModelChecker::check_current_action(ModelAction *curr)
1236 bool second_part_of_rmw = curr->is_rmwc() || curr->is_rmw();
1238 if (!check_action_enabled(curr)) {
1239 /* Make the execution look like we chose to run this action
1240 * much later, when a lock/join can succeed */
1241 get_thread(curr)->set_pending(curr);
1242 scheduler->sleep(get_thread(curr));
1246 bool newly_explored = initialize_curr_action(&curr);
1252 wake_up_sleeping_actions(curr);
1254 /* Add the action to lists before any other model-checking tasks */
1255 if (!second_part_of_rmw)
1256 add_action_to_lists(curr);
1258 /* Build may_read_from set for newly-created actions */
1259 if (newly_explored && curr->is_read())
1260 build_reads_from_past(curr);
1262 /* Initialize work_queue with the "current action" work */
1263 work_queue_t work_queue(1, CheckCurrWorkEntry(curr));
1264 while (!work_queue.empty() && !has_asserted()) {
1265 WorkQueueEntry work = work_queue.front();
1266 work_queue.pop_front();
1268 switch (work.type) {
1269 case WORK_CHECK_CURR_ACTION: {
1270 ModelAction *act = work.action;
1271 bool update = false; /* update this location's release seq's */
1272 bool update_all = false; /* update all release seq's */
1274 if (process_thread_action(curr))
1277 if (act->is_read() && process_read(act, second_part_of_rmw))
1280 if (act->is_write() && process_write(act))
1283 if (act->is_fence() && process_fence(act))
1286 if (act->is_mutex_op() && process_mutex(act))
1289 if (act->is_relseq_fixup())
1290 process_relseq_fixup(curr, &work_queue);
1293 work_queue.push_back(CheckRelSeqWorkEntry(NULL));
1295 work_queue.push_back(CheckRelSeqWorkEntry(act->get_location()));
1298 case WORK_CHECK_RELEASE_SEQ:
1299 resolve_release_sequences(work.location, &work_queue);
1301 case WORK_CHECK_MO_EDGES: {
1302 /** @todo Complete verification of work_queue */
1303 ModelAction *act = work.action;
1304 bool updated = false;
1306 if (act->is_read()) {
1307 const ModelAction *rf = act->get_reads_from();
1308 if (rf != NULL && r_modification_order(act, rf))
1311 if (act->is_write()) {
1312 if (w_modification_order(act))
1315 mo_graph->commitChanges();
1318 work_queue.push_back(CheckRelSeqWorkEntry(act->get_location()));
1327 check_curr_backtracking(curr);
1328 set_backtracking(curr);
1332 void ModelChecker::check_curr_backtracking(ModelAction *curr)
1334 Node *currnode = curr->get_node();
1335 Node *parnode = currnode->get_parent();
1337 if ((parnode && !parnode->backtrack_empty()) ||
1338 !currnode->misc_empty() ||
1339 !currnode->read_from_empty() ||
1340 !currnode->future_value_empty() ||
1341 !currnode->promise_empty() ||
1342 !currnode->relseq_break_empty()) {
1343 set_latest_backtrack(curr);
1347 bool ModelChecker::promises_expired() const
1349 for (unsigned int i = 0; i < promises->size(); i++) {
1350 Promise *promise = (*promises)[i];
1351 if (promise->get_expiration() < priv->used_sequence_numbers)
1358 * This is the strongest feasibility check available.
1359 * @return whether the current trace (partial or complete) must be a prefix of
1362 bool ModelChecker::isfeasibleprefix() const
1364 return pending_rel_seqs->size() == 0 && is_feasible_prefix_ignore_relseq();
1368 * Print disagnostic information about an infeasible execution
1369 * @param prefix A string to prefix the output with; if NULL, then a default
1370 * message prefix will be provided
1372 void ModelChecker::print_infeasibility(const char *prefix) const
1376 if (mo_graph->checkForRMWViolation())
1377 ptr += sprintf(ptr, "[RMW atomicity]");
1378 if (mo_graph->checkForCycles())
1379 ptr += sprintf(ptr, "[mo cycle]");
1380 if (priv->failed_promise)
1381 ptr += sprintf(ptr, "[failed promise]");
1382 if (priv->too_many_reads)
1383 ptr += sprintf(ptr, "[too many reads]");
1384 if (priv->bad_synchronization)
1385 ptr += sprintf(ptr, "[bad sw ordering]");
1386 if (promises_expired())
1387 ptr += sprintf(ptr, "[promise expired]");
1388 if (promises->size() != 0)
1389 ptr += sprintf(ptr, "[unresolved promise]");
1391 model_print("%s: %s\n", prefix ? prefix : "Infeasible", buf);
1395 * Returns whether the current completed trace is feasible, except for pending
1396 * release sequences.
1398 bool ModelChecker::is_feasible_prefix_ignore_relseq() const
1400 return !is_infeasible() && promises->size() == 0;
1404 * Check if the current partial trace is infeasible. Does not check any
1405 * end-of-execution flags, which might rule out the execution. Thus, this is
1406 * useful only for ruling an execution as infeasible.
1407 * @return whether the current partial trace is infeasible.
1409 bool ModelChecker::is_infeasible() const
1411 return mo_graph->checkForRMWViolation() || is_infeasible_ignoreRMW();
1415 * Check If the current partial trace is infeasible, while ignoring
1416 * infeasibility related to 2 RMW's reading from the same store. It does not
1417 * check end-of-execution feasibility.
1418 * @see ModelChecker::is_infeasible
1419 * @return whether the current partial trace is infeasible, ignoring multiple
1420 * RMWs reading from the same store.
1422 bool ModelChecker::is_infeasible_ignoreRMW() const
1424 return mo_graph->checkForCycles() || priv->failed_promise ||
1425 priv->too_many_reads || priv->bad_synchronization ||
1429 /** Close out a RMWR by converting previous RMWR into a RMW or READ. */
1430 ModelAction * ModelChecker::process_rmw(ModelAction *act) {
1431 ModelAction *lastread = get_last_action(act->get_tid());
1432 lastread->process_rmw(act);
1433 if (act->is_rmw() && lastread->get_reads_from() != NULL) {
1434 mo_graph->addRMWEdge(lastread->get_reads_from(), lastread);
1435 mo_graph->commitChanges();
1441 * Checks whether a thread has read from the same write for too many times
1442 * without seeing the effects of a later write.
1445 * 1) there must a different write that we could read from that would satisfy the modification order,
1446 * 2) we must have read from the same value in excess of maxreads times, and
1447 * 3) that other write must have been in the reads_from set for maxreads times.
1449 * If so, we decide that the execution is no longer feasible.
1451 void ModelChecker::check_recency(ModelAction *curr, const ModelAction *rf)
1453 if (params.maxreads != 0) {
1454 if (curr->get_node()->get_read_from_size() <= 1)
1456 //Must make sure that execution is currently feasible... We could
1457 //accidentally clear by rolling back
1458 if (is_infeasible())
1460 std::vector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1461 int tid = id_to_int(curr->get_tid());
1464 if ((int)thrd_lists->size() <= tid)
1466 action_list_t *list = &(*thrd_lists)[tid];
1468 action_list_t::reverse_iterator rit = list->rbegin();
1469 /* Skip past curr */
1470 for (; (*rit) != curr; rit++)
1472 /* go past curr now */
1475 action_list_t::reverse_iterator ritcopy = rit;
1476 //See if we have enough reads from the same value
1478 for (; count < params.maxreads; rit++, count++) {
1479 if (rit == list->rend())
1481 ModelAction *act = *rit;
1482 if (!act->is_read())
1485 if (act->get_reads_from() != rf)
1487 if (act->get_node()->get_read_from_size() <= 1)
1490 for (int i = 0; i < curr->get_node()->get_read_from_size(); i++) {
1492 const ModelAction *write = curr->get_node()->get_read_from_at(i);
1494 /* Need a different write */
1498 /* Test to see whether this is a feasible write to read from */
1499 mo_graph->startChanges();
1500 r_modification_order(curr, write);
1501 bool feasiblereadfrom = !is_infeasible();
1502 mo_graph->rollbackChanges();
1504 if (!feasiblereadfrom)
1508 bool feasiblewrite = true;
1509 //new we need to see if this write works for everyone
1511 for (int loop = count; loop > 0; loop--, rit++) {
1512 ModelAction *act = *rit;
1513 bool foundvalue = false;
1514 for (int j = 0; j < act->get_node()->get_read_from_size(); j++) {
1515 if (act->get_node()->get_read_from_at(j) == write) {
1521 feasiblewrite = false;
1525 if (feasiblewrite) {
1526 priv->too_many_reads = true;
1534 * Updates the mo_graph with the constraints imposed from the current
1537 * Basic idea is the following: Go through each other thread and find
1538 * the lastest action that happened before our read. Two cases:
1540 * (1) The action is a write => that write must either occur before
1541 * the write we read from or be the write we read from.
1543 * (2) The action is a read => the write that that action read from
1544 * must occur before the write we read from or be the same write.
1546 * @param curr The current action. Must be a read.
1547 * @param rf The action that curr reads from. Must be a write.
1548 * @return True if modification order edges were added; false otherwise
1550 bool ModelChecker::r_modification_order(ModelAction *curr, const ModelAction *rf)
1552 std::vector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1555 ASSERT(curr->is_read());
1557 /* Last SC fence in the current thread */
1558 ModelAction *last_sc_fence_local = get_last_seq_cst_fence(curr->get_tid(), NULL);
1560 /* Iterate over all threads */
1561 for (i = 0; i < thrd_lists->size(); i++) {
1562 /* Last SC fence in thread i */
1563 ModelAction *last_sc_fence_thread_local = NULL;
1564 if (int_to_id((int)i) != curr->get_tid())
1565 last_sc_fence_thread_local = get_last_seq_cst_fence(int_to_id(i), NULL);
1567 /* Last SC fence in thread i, before last SC fence in current thread */
1568 ModelAction *last_sc_fence_thread_before = NULL;
1569 if (last_sc_fence_local)
1570 last_sc_fence_thread_before = get_last_seq_cst_fence(int_to_id(i), last_sc_fence_local);
1572 /* Iterate over actions in thread, starting from most recent */
1573 action_list_t *list = &(*thrd_lists)[i];
1574 action_list_t::reverse_iterator rit;
1575 for (rit = list->rbegin(); rit != list->rend(); rit++) {
1576 ModelAction *act = *rit;
1578 if (act->is_write() && act != rf && act != curr) {
1579 /* C++, Section 29.3 statement 5 */
1580 if (curr->is_seqcst() && last_sc_fence_thread_local &&
1581 *act < *last_sc_fence_thread_local) {
1582 mo_graph->addEdge(act, rf);
1586 /* C++, Section 29.3 statement 4 */
1587 else if (act->is_seqcst() && last_sc_fence_local &&
1588 *act < *last_sc_fence_local) {
1589 mo_graph->addEdge(act, rf);
1593 /* C++, Section 29.3 statement 6 */
1594 else if (last_sc_fence_thread_before &&
1595 *act < *last_sc_fence_thread_before) {
1596 mo_graph->addEdge(act, rf);
1603 * Include at most one act per-thread that "happens
1604 * before" curr. Don't consider reflexively.
1606 if (act->happens_before(curr) && act != curr) {
1607 if (act->is_write()) {
1609 mo_graph->addEdge(act, rf);
1613 const ModelAction *prevreadfrom = act->get_reads_from();
1614 //if the previous read is unresolved, keep going...
1615 if (prevreadfrom == NULL)
1618 if (rf != prevreadfrom) {
1619 mo_graph->addEdge(prevreadfrom, rf);
1631 /** This method fixes up the modification order when we resolve a
1632 * promises. The basic problem is that actions that occur after the
1633 * read curr could not property add items to the modification order
1636 * So for each thread, we find the earliest item that happens after
1637 * the read curr. This is the item we have to fix up with additional
1638 * constraints. If that action is write, we add a MO edge between
1639 * the Action rf and that action. If the action is a read, we add a
1640 * MO edge between the Action rf, and whatever the read accessed.
1642 * @param curr is the read ModelAction that we are fixing up MO edges for.
1643 * @param rf is the write ModelAction that curr reads from.
1646 void ModelChecker::post_r_modification_order(ModelAction *curr, const ModelAction *rf)
1648 std::vector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1650 ASSERT(curr->is_read());
1652 /* Iterate over all threads */
1653 for (i = 0; i < thrd_lists->size(); i++) {
1654 /* Iterate over actions in thread, starting from most recent */
1655 action_list_t *list = &(*thrd_lists)[i];
1656 action_list_t::reverse_iterator rit;
1657 ModelAction *lastact = NULL;
1659 /* Find last action that happens after curr that is either not curr or a rmw */
1660 for (rit = list->rbegin(); rit != list->rend(); rit++) {
1661 ModelAction *act = *rit;
1662 if (curr->happens_before(act) && (curr != act || curr->is_rmw())) {
1668 /* Include at most one act per-thread that "happens before" curr */
1669 if (lastact != NULL) {
1670 if (lastact == curr) {
1671 //Case 1: The resolved read is a RMW, and we need to make sure
1672 //that the write portion of the RMW mod order after rf
1674 mo_graph->addEdge(rf, lastact);
1675 } else if (lastact->is_read()) {
1676 //Case 2: The resolved read is a normal read and the next
1677 //operation is a read, and we need to make sure the value read
1678 //is mod ordered after rf
1680 const ModelAction *postreadfrom = lastact->get_reads_from();
1681 if (postreadfrom != NULL && rf != postreadfrom)
1682 mo_graph->addEdge(rf, postreadfrom);
1684 //Case 3: The resolved read is a normal read and the next
1685 //operation is a write, and we need to make sure that the
1686 //write is mod ordered after rf
1688 mo_graph->addEdge(rf, lastact);
1696 * Updates the mo_graph with the constraints imposed from the current write.
1698 * Basic idea is the following: Go through each other thread and find
1699 * the lastest action that happened before our write. Two cases:
1701 * (1) The action is a write => that write must occur before
1704 * (2) The action is a read => the write that that action read from
1705 * must occur before the current write.
1707 * This method also handles two other issues:
1709 * (I) Sequential Consistency: Making sure that if the current write is
1710 * seq_cst, that it occurs after the previous seq_cst write.
1712 * (II) Sending the write back to non-synchronizing reads.
1714 * @param curr The current action. Must be a write.
1715 * @return True if modification order edges were added; false otherwise
1717 bool ModelChecker::w_modification_order(ModelAction *curr)
1719 std::vector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1722 ASSERT(curr->is_write());
1724 if (curr->is_seqcst()) {
1725 /* We have to at least see the last sequentially consistent write,
1726 so we are initialized. */
1727 ModelAction *last_seq_cst = get_last_seq_cst_write(curr);
1728 if (last_seq_cst != NULL) {
1729 mo_graph->addEdge(last_seq_cst, curr);
1734 /* Last SC fence in the current thread */
1735 ModelAction *last_sc_fence_local = get_last_seq_cst_fence(curr->get_tid(), NULL);
1737 /* Iterate over all threads */
1738 for (i = 0; i < thrd_lists->size(); i++) {
1739 /* Last SC fence in thread i, before last SC fence in current thread */
1740 ModelAction *last_sc_fence_thread_before = NULL;
1741 if (last_sc_fence_local && int_to_id((int)i) != curr->get_tid())
1742 last_sc_fence_thread_before = get_last_seq_cst_fence(int_to_id(i), last_sc_fence_local);
1744 /* Iterate over actions in thread, starting from most recent */
1745 action_list_t *list = &(*thrd_lists)[i];
1746 action_list_t::reverse_iterator rit;
1747 for (rit = list->rbegin(); rit != list->rend(); rit++) {
1748 ModelAction *act = *rit;
1751 * 1) If RMW and it actually read from something, then we
1752 * already have all relevant edges, so just skip to next
1755 * 2) If RMW and it didn't read from anything, we should
1756 * whatever edge we can get to speed up convergence.
1758 * 3) If normal write, we need to look at earlier actions, so
1759 * continue processing list.
1761 if (curr->is_rmw()) {
1762 if (curr->get_reads_from() != NULL)
1770 /* C++, Section 29.3 statement 7 */
1771 if (last_sc_fence_thread_before && act->is_write() &&
1772 *act < *last_sc_fence_thread_before) {
1773 mo_graph->addEdge(act, curr);
1779 * Include at most one act per-thread that "happens
1782 if (act->happens_before(curr)) {
1784 * Note: if act is RMW, just add edge:
1786 * The following edge should be handled elsewhere:
1787 * readfrom(act) --mo--> act
1789 if (act->is_write())
1790 mo_graph->addEdge(act, curr);
1791 else if (act->is_read()) {
1792 //if previous read accessed a null, just keep going
1793 if (act->get_reads_from() == NULL)
1795 mo_graph->addEdge(act->get_reads_from(), curr);
1799 } else if (act->is_read() && !act->could_synchronize_with(curr) &&
1800 !act->same_thread(curr)) {
1801 /* We have an action that:
1802 (1) did not happen before us
1803 (2) is a read and we are a write
1804 (3) cannot synchronize with us
1805 (4) is in a different thread
1807 that read could potentially read from our write. Note that
1808 these checks are overly conservative at this point, we'll
1809 do more checks before actually removing the
1813 if (thin_air_constraint_may_allow(curr, act)) {
1814 if (!is_infeasible() ||
1815 (curr->is_rmw() && act->is_rmw() && curr->get_reads_from() == act->get_reads_from() && !is_infeasible_ignoreRMW())) {
1816 futurevalues->push_back(PendingFutureValue(curr, act));
1826 /** Arbitrary reads from the future are not allowed. Section 29.3
1827 * part 9 places some constraints. This method checks one result of constraint
1828 * constraint. Others require compiler support. */
1829 bool ModelChecker::thin_air_constraint_may_allow(const ModelAction *writer, const ModelAction *reader)
1831 if (!writer->is_rmw())
1834 if (!reader->is_rmw())
1837 for (const ModelAction *search = writer->get_reads_from(); search != NULL; search = search->get_reads_from()) {
1838 if (search == reader)
1840 if (search->get_tid() == reader->get_tid() &&
1841 search->happens_before(reader))
1849 * Arbitrary reads from the future are not allowed. Section 29.3 part 9 places
1850 * some constraints. This method checks one the following constraint (others
1851 * require compiler support):
1853 * If X --hb-> Y --mo-> Z, then X should not read from Z.
1855 bool ModelChecker::mo_may_allow(const ModelAction *writer, const ModelAction *reader)
1857 std::vector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, reader->get_location());
1859 /* Iterate over all threads */
1860 for (i = 0; i < thrd_lists->size(); i++) {
1861 const ModelAction *write_after_read = NULL;
1863 /* Iterate over actions in thread, starting from most recent */
1864 action_list_t *list = &(*thrd_lists)[i];
1865 action_list_t::reverse_iterator rit;
1866 for (rit = list->rbegin(); rit != list->rend(); rit++) {
1867 ModelAction *act = *rit;
1869 /* Don't disallow due to act == reader */
1870 if (!reader->happens_before(act) || reader == act)
1872 else if (act->is_write())
1873 write_after_read = act;
1874 else if (act->is_read() && act->get_reads_from() != NULL)
1875 write_after_read = act->get_reads_from();
1878 if (write_after_read && write_after_read != writer && mo_graph->checkReachable(write_after_read, writer))
1885 * Finds the head(s) of the release sequence(s) containing a given ModelAction.
1886 * The ModelAction under consideration is expected to be taking part in
1887 * release/acquire synchronization as an object of the "reads from" relation.
1888 * Note that this can only provide release sequence support for RMW chains
1889 * which do not read from the future, as those actions cannot be traced until
1890 * their "promise" is fulfilled. Similarly, we may not even establish the
1891 * presence of a release sequence with certainty, as some modification order
1892 * constraints may be decided further in the future. Thus, this function
1893 * "returns" two pieces of data: a pass-by-reference vector of @a release_heads
1894 * and a boolean representing certainty.
1896 * @param rf The action that might be part of a release sequence. Must be a
1898 * @param release_heads A pass-by-reference style return parameter. After
1899 * execution of this function, release_heads will contain the heads of all the
1900 * relevant release sequences, if any exists with certainty
1901 * @param pending A pass-by-reference style return parameter which is only used
1902 * when returning false (i.e., uncertain). Returns most information regarding
1903 * an uncertain release sequence, including any write operations that might
1904 * break the sequence.
1905 * @return true, if the ModelChecker is certain that release_heads is complete;
1908 bool ModelChecker::release_seq_heads(const ModelAction *rf,
1909 rel_heads_list_t *release_heads,
1910 struct release_seq *pending) const
1912 /* Only check for release sequences if there are no cycles */
1913 if (mo_graph->checkForCycles())
1917 ASSERT(rf->is_write());
1919 if (rf->is_release())
1920 release_heads->push_back(rf);
1921 else if (rf->get_last_fence_release())
1922 release_heads->push_back(rf->get_last_fence_release());
1924 break; /* End of RMW chain */
1926 /** @todo Need to be smarter here... In the linux lock
1927 * example, this will run to the beginning of the program for
1929 /** @todo The way to be smarter here is to keep going until 1
1930 * thread has a release preceded by an acquire and you've seen
1933 /* acq_rel RMW is a sufficient stopping condition */
1934 if (rf->is_acquire() && rf->is_release())
1935 return true; /* complete */
1937 rf = rf->get_reads_from();
1940 /* read from future: need to settle this later */
1942 return false; /* incomplete */
1945 if (rf->is_release())
1946 return true; /* complete */
1948 /* else relaxed write
1949 * - check for fence-release in the same thread (29.8, stmt. 3)
1950 * - check modification order for contiguous subsequence
1951 * -> rf must be same thread as release */
1953 const ModelAction *fence_release = rf->get_last_fence_release();
1954 /* Synchronize with a fence-release unconditionally; we don't need to
1955 * find any more "contiguous subsequence..." for it */
1957 release_heads->push_back(fence_release);
1959 int tid = id_to_int(rf->get_tid());
1960 std::vector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, rf->get_location());
1961 action_list_t *list = &(*thrd_lists)[tid];
1962 action_list_t::const_reverse_iterator rit;
1964 /* Find rf in the thread list */
1965 rit = std::find(list->rbegin(), list->rend(), rf);
1966 ASSERT(rit != list->rend());
1968 /* Find the last {write,fence}-release */
1969 for (; rit != list->rend(); rit++) {
1970 if (fence_release && *(*rit) < *fence_release)
1972 if ((*rit)->is_release())
1975 if (rit == list->rend()) {
1976 /* No write-release in this thread */
1977 return true; /* complete */
1978 } else if (fence_release && *(*rit) < *fence_release) {
1979 /* The fence-release is more recent (and so, "stronger") than
1980 * the most recent write-release */
1981 return true; /* complete */
1982 } /* else, need to establish contiguous release sequence */
1983 ModelAction *release = *rit;
1985 ASSERT(rf->same_thread(release));
1987 pending->writes.clear();
1989 bool certain = true;
1990 for (unsigned int i = 0; i < thrd_lists->size(); i++) {
1991 if (id_to_int(rf->get_tid()) == (int)i)
1993 list = &(*thrd_lists)[i];
1995 /* Can we ensure no future writes from this thread may break
1996 * the release seq? */
1997 bool future_ordered = false;
1999 ModelAction *last = get_last_action(int_to_id(i));
2000 Thread *th = get_thread(int_to_id(i));
2001 if ((last && rf->happens_before(last)) ||
2004 future_ordered = true;
2006 ASSERT(!th->is_model_thread() || future_ordered);
2008 for (rit = list->rbegin(); rit != list->rend(); rit++) {
2009 const ModelAction *act = *rit;
2010 /* Reach synchronization -> this thread is complete */
2011 if (act->happens_before(release))
2013 if (rf->happens_before(act)) {
2014 future_ordered = true;
2018 /* Only non-RMW writes can break release sequences */
2019 if (!act->is_write() || act->is_rmw())
2022 /* Check modification order */
2023 if (mo_graph->checkReachable(rf, act)) {
2024 /* rf --mo--> act */
2025 future_ordered = true;
2028 if (mo_graph->checkReachable(act, release))
2029 /* act --mo--> release */
2031 if (mo_graph->checkReachable(release, act) &&
2032 mo_graph->checkReachable(act, rf)) {
2033 /* release --mo-> act --mo--> rf */
2034 return true; /* complete */
2036 /* act may break release sequence */
2037 pending->writes.push_back(act);
2040 if (!future_ordered)
2041 certain = false; /* This thread is uncertain */
2045 release_heads->push_back(release);
2046 pending->writes.clear();
2048 pending->release = release;
2055 * An interface for getting the release sequence head(s) with which a
2056 * given ModelAction must synchronize. This function only returns a non-empty
2057 * result when it can locate a release sequence head with certainty. Otherwise,
2058 * it may mark the internal state of the ModelChecker so that it will handle
2059 * the release sequence at a later time, causing @a acquire to update its
2060 * synchronization at some later point in execution.
2062 * @param acquire The 'acquire' action that may synchronize with a release
2064 * @param read The read action that may read from a release sequence; this may
2065 * be the same as acquire, or else an earlier action in the same thread (i.e.,
2066 * when 'acquire' is a fence-acquire)
2067 * @param release_heads A pass-by-reference return parameter. Will be filled
2068 * with the head(s) of the release sequence(s), if they exists with certainty.
2069 * @see ModelChecker::release_seq_heads
2071 void ModelChecker::get_release_seq_heads(ModelAction *acquire,
2072 ModelAction *read, rel_heads_list_t *release_heads)
2074 const ModelAction *rf = read->get_reads_from();
2075 struct release_seq *sequence = (struct release_seq *)snapshot_calloc(1, sizeof(struct release_seq));
2076 sequence->acquire = acquire;
2077 sequence->read = read;
2079 if (!release_seq_heads(rf, release_heads, sequence)) {
2080 /* add act to 'lazy checking' list */
2081 pending_rel_seqs->push_back(sequence);
2083 snapshot_free(sequence);
2088 * Attempt to resolve all stashed operations that might synchronize with a
2089 * release sequence for a given location. This implements the "lazy" portion of
2090 * determining whether or not a release sequence was contiguous, since not all
2091 * modification order information is present at the time an action occurs.
2093 * @param location The location/object that should be checked for release
2094 * sequence resolutions. A NULL value means to check all locations.
2095 * @param work_queue The work queue to which to add work items as they are
2097 * @return True if any updates occurred (new synchronization, new mo_graph
2100 bool ModelChecker::resolve_release_sequences(void *location, work_queue_t *work_queue)
2102 bool updated = false;
2103 std::vector< struct release_seq *, SnapshotAlloc<struct release_seq *> >::iterator it = pending_rel_seqs->begin();
2104 while (it != pending_rel_seqs->end()) {
2105 struct release_seq *pending = *it;
2106 ModelAction *acquire = pending->acquire;
2107 const ModelAction *read = pending->read;
2109 /* Only resolve sequences on the given location, if provided */
2110 if (location && read->get_location() != location) {
2115 const ModelAction *rf = read->get_reads_from();
2116 rel_heads_list_t release_heads;
2118 complete = release_seq_heads(rf, &release_heads, pending);
2119 for (unsigned int i = 0; i < release_heads.size(); i++) {
2120 if (!acquire->has_synchronized_with(release_heads[i])) {
2121 if (acquire->synchronize_with(release_heads[i]))
2124 set_bad_synchronization();
2129 /* Re-check all pending release sequences */
2130 work_queue->push_back(CheckRelSeqWorkEntry(NULL));
2131 /* Re-check read-acquire for mo_graph edges */
2132 if (acquire->is_read())
2133 work_queue->push_back(MOEdgeWorkEntry(acquire));
2135 /* propagate synchronization to later actions */
2136 action_list_t::reverse_iterator rit = action_trace->rbegin();
2137 for (; (*rit) != acquire; rit++) {
2138 ModelAction *propagate = *rit;
2139 if (acquire->happens_before(propagate)) {
2140 propagate->synchronize_with(acquire);
2141 /* Re-check 'propagate' for mo_graph edges */
2142 work_queue->push_back(MOEdgeWorkEntry(propagate));
2147 it = pending_rel_seqs->erase(it);
2148 snapshot_free(pending);
2154 // If we resolved promises or data races, see if we have realized a data race.
2161 * Performs various bookkeeping operations for the current ModelAction. For
2162 * instance, adds action to the per-object, per-thread action vector and to the
2163 * action trace list of all thread actions.
2165 * @param act is the ModelAction to add.
2167 void ModelChecker::add_action_to_lists(ModelAction *act)
2169 int tid = id_to_int(act->get_tid());
2170 ModelAction *uninit = NULL;
2172 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
2173 if (list->empty() && act->is_atomic_var()) {
2174 uninit = new_uninitialized_action(act->get_location());
2175 uninit_id = id_to_int(uninit->get_tid());
2176 list->push_back(uninit);
2178 list->push_back(act);
2180 action_trace->push_back(act);
2182 action_trace->push_front(uninit);
2184 std::vector<action_list_t> *vec = get_safe_ptr_vect_action(obj_thrd_map, act->get_location());
2185 if (tid >= (int)vec->size())
2186 vec->resize(priv->next_thread_id);
2187 (*vec)[tid].push_back(act);
2189 (*vec)[uninit_id].push_front(uninit);
2191 if ((int)thrd_last_action->size() <= tid)
2192 thrd_last_action->resize(get_num_threads());
2193 (*thrd_last_action)[tid] = act;
2195 (*thrd_last_action)[uninit_id] = uninit;
2197 if (act->is_fence() && act->is_release()) {
2198 if ((int)thrd_last_fence_release->size() <= tid)
2199 thrd_last_fence_release->resize(get_num_threads());
2200 (*thrd_last_fence_release)[tid] = act;
2203 if (act->is_wait()) {
2204 void *mutex_loc = (void *) act->get_value();
2205 get_safe_ptr_action(obj_map, mutex_loc)->push_back(act);
2207 std::vector<action_list_t> *vec = get_safe_ptr_vect_action(obj_thrd_map, mutex_loc);
2208 if (tid >= (int)vec->size())
2209 vec->resize(priv->next_thread_id);
2210 (*vec)[tid].push_back(act);
2215 * @brief Get the last action performed by a particular Thread
2216 * @param tid The thread ID of the Thread in question
2217 * @return The last action in the thread
2219 ModelAction * ModelChecker::get_last_action(thread_id_t tid) const
2221 int threadid = id_to_int(tid);
2222 if (threadid < (int)thrd_last_action->size())
2223 return (*thrd_last_action)[id_to_int(tid)];
2229 * @brief Get the last fence release performed by a particular Thread
2230 * @param tid The thread ID of the Thread in question
2231 * @return The last fence release in the thread, if one exists; NULL otherwise
2233 ModelAction * ModelChecker::get_last_fence_release(thread_id_t tid) const
2235 int threadid = id_to_int(tid);
2236 if (threadid < (int)thrd_last_fence_release->size())
2237 return (*thrd_last_fence_release)[id_to_int(tid)];
2243 * Gets the last memory_order_seq_cst write (in the total global sequence)
2244 * performed on a particular object (i.e., memory location), not including the
2246 * @param curr The current ModelAction; also denotes the object location to
2248 * @return The last seq_cst write
2250 ModelAction * ModelChecker::get_last_seq_cst_write(ModelAction *curr) const
2252 void *location = curr->get_location();
2253 action_list_t *list = get_safe_ptr_action(obj_map, location);
2254 /* Find: max({i in dom(S) | seq_cst(t_i) && isWrite(t_i) && samevar(t_i, t)}) */
2255 action_list_t::reverse_iterator rit;
2256 for (rit = list->rbegin(); rit != list->rend(); rit++)
2257 if ((*rit)->is_write() && (*rit)->is_seqcst() && (*rit) != curr)
2263 * Gets the last memory_order_seq_cst fence (in the total global sequence)
2264 * performed in a particular thread, prior to a particular fence.
2265 * @param tid The ID of the thread to check
2266 * @param before_fence The fence from which to begin the search; if NULL, then
2267 * search for the most recent fence in the thread.
2268 * @return The last prior seq_cst fence in the thread, if exists; otherwise, NULL
2270 ModelAction * ModelChecker::get_last_seq_cst_fence(thread_id_t tid, const ModelAction *before_fence) const
2272 /* All fences should have NULL location */
2273 action_list_t *list = get_safe_ptr_action(obj_map, NULL);
2274 action_list_t::reverse_iterator rit = list->rbegin();
2277 for (; rit != list->rend(); rit++)
2278 if (*rit == before_fence)
2281 ASSERT(*rit == before_fence);
2285 for (; rit != list->rend(); rit++)
2286 if ((*rit)->is_fence() && (tid == (*rit)->get_tid()) && (*rit)->is_seqcst())
2292 * Gets the last unlock operation performed on a particular mutex (i.e., memory
2293 * location). This function identifies the mutex according to the current
2294 * action, which is presumed to perform on the same mutex.
2295 * @param curr The current ModelAction; also denotes the object location to
2297 * @return The last unlock operation
2299 ModelAction * ModelChecker::get_last_unlock(ModelAction *curr) const
2301 void *location = curr->get_location();
2302 action_list_t *list = get_safe_ptr_action(obj_map, location);
2303 /* Find: max({i in dom(S) | isUnlock(t_i) && samevar(t_i, t)}) */
2304 action_list_t::reverse_iterator rit;
2305 for (rit = list->rbegin(); rit != list->rend(); rit++)
2306 if ((*rit)->is_unlock() || (*rit)->is_wait())
2311 ModelAction * ModelChecker::get_parent_action(thread_id_t tid) const
2313 ModelAction *parent = get_last_action(tid);
2315 parent = get_thread(tid)->get_creation();
2320 * Returns the clock vector for a given thread.
2321 * @param tid The thread whose clock vector we want
2322 * @return Desired clock vector
2324 ClockVector * ModelChecker::get_cv(thread_id_t tid) const
2326 return get_parent_action(tid)->get_cv();
2330 * Resolve a set of Promises with a current write. The set is provided in the
2331 * Node corresponding to @a write.
2332 * @param write The ModelAction that is fulfilling Promises
2333 * @return True if promises were resolved; false otherwise
2335 bool ModelChecker::resolve_promises(ModelAction *write)
2337 bool resolved = false;
2338 std::vector< ModelAction *, ModelAlloc<ModelAction *> > actions_to_check;
2340 for (unsigned int i = 0, promise_index = 0; promise_index < promises->size(); i++) {
2341 Promise *promise = (*promises)[promise_index];
2342 if (write->get_node()->get_promise(i)) {
2343 ModelAction *read = promise->get_action();
2344 if (read->is_rmw()) {
2345 mo_graph->addRMWEdge(write, read);
2347 read_from(read, write);
2348 //First fix up the modification order for actions that happened
2350 r_modification_order(read, write);
2351 //Next fix up the modification order for actions that happened
2353 post_r_modification_order(read, write);
2354 //Make sure the promise's value matches the write's value
2355 ASSERT(promise->get_value() == write->get_value());
2358 promises->erase(promises->begin() + promise_index);
2359 actions_to_check.push_back(read);
2366 //Check whether reading these writes has made threads unable to
2369 for (unsigned int i = 0; i < actions_to_check.size(); i++) {
2370 ModelAction *read=actions_to_check[i];
2371 mo_check_promises(read->get_tid(), write, read);
2378 * Compute the set of promises that could potentially be satisfied by this
2379 * action. Note that the set computation actually appears in the Node, not in
2381 * @param curr The ModelAction that may satisfy promises
2383 void ModelChecker::compute_promises(ModelAction *curr)
2385 for (unsigned int i = 0; i < promises->size(); i++) {
2386 Promise *promise = (*promises)[i];
2387 const ModelAction *act = promise->get_action();
2388 if (!act->happens_before(curr) &&
2390 !act->could_synchronize_with(curr) &&
2391 !act->same_thread(curr) &&
2392 act->get_location() == curr->get_location() &&
2393 promise->get_value() == curr->get_value()) {
2394 curr->get_node()->set_promise(i, act->is_rmw());
2399 /** Checks promises in response to change in ClockVector Threads. */
2400 void ModelChecker::check_promises(thread_id_t tid, ClockVector *old_cv, ClockVector *merge_cv)
2402 for (unsigned int i = 0; i < promises->size(); i++) {
2403 Promise *promise = (*promises)[i];
2404 const ModelAction *act = promise->get_action();
2405 if ((old_cv == NULL || !old_cv->synchronized_since(act)) &&
2406 merge_cv->synchronized_since(act)) {
2407 if (promise->eliminate_thread(tid)) {
2408 //Promise has failed
2409 priv->failed_promise = true;
2416 void ModelChecker::check_promises_thread_disabled()
2418 for (unsigned int i = 0; i < promises->size(); i++) {
2419 Promise *promise = (*promises)[i];
2420 if (promise->has_failed()) {
2421 priv->failed_promise = true;
2428 * @brief Checks promises in response to addition to modification order for
2433 * pthread is the thread that performed the read that created the promise
2435 * pread is the read that created the promise
2437 * pwrite is either the first write to same location as pread by
2438 * pthread that is sequenced after pread or the write read by the
2439 * first read to the same location as pread by pthread that is
2440 * sequenced after pread.
2442 * 1. If tid=pthread, then we check what other threads are reachable
2443 * through the mod order starting with pwrite. Those threads cannot
2444 * perform a write that will resolve the promise due to modification
2445 * order constraints.
2447 * 2. If the tid is not pthread, we check whether pwrite can reach the
2448 * action write through the modification order. If so, that thread
2449 * cannot perform a future write that will resolve the promise due to
2450 * modificatin order constraints.
2452 * @param tid The thread that either read from the model action write, or
2453 * actually did the model action write.
2455 * @param write The ModelAction representing the relevant write.
2456 * @param read The ModelAction that reads a promised write, or NULL otherwise.
2458 void ModelChecker::mo_check_promises(thread_id_t tid, const ModelAction *write, const ModelAction *read)
2460 void *location = write->get_location();
2461 for (unsigned int i = 0; i < promises->size(); i++) {
2462 Promise *promise = (*promises)[i];
2463 const ModelAction *act = promise->get_action();
2465 // Is this promise on the same location?
2466 if (act->get_location() != location)
2469 // same thread as the promise
2470 if (act->get_tid() == tid) {
2471 // make sure that the reader of this write happens after the promise
2472 if ((read == NULL) || (promise->get_action()->happens_before(read))) {
2473 // do we have a pwrite for the promise, if not, set it
2474 if (promise->get_write() == NULL) {
2475 promise->set_write(write);
2476 // The pwrite cannot happen before the promise
2477 if (write->happens_before(act) && (write != act)) {
2478 priv->failed_promise = true;
2483 if (mo_graph->checkPromise(write, promise)) {
2484 priv->failed_promise = true;
2490 // Don't do any lookups twice for the same thread
2491 if (!promise->thread_is_available(tid))
2494 if (promise->get_write() && mo_graph->checkReachable(promise->get_write(), write)) {
2495 if (promise->eliminate_thread(tid)) {
2496 priv->failed_promise = true;
2504 * Compute the set of writes that may break the current pending release
2505 * sequence. This information is extracted from previou release sequence
2508 * @param curr The current ModelAction. Must be a release sequence fixup
2511 void ModelChecker::compute_relseq_breakwrites(ModelAction *curr)
2513 if (pending_rel_seqs->empty())
2516 struct release_seq *pending = pending_rel_seqs->back();
2517 for (unsigned int i = 0; i < pending->writes.size(); i++) {
2518 const ModelAction *write = pending->writes[i];
2519 curr->get_node()->add_relseq_break(write);
2522 /* NULL means don't break the sequence; just synchronize */
2523 curr->get_node()->add_relseq_break(NULL);
2527 * Build up an initial set of all past writes that this 'read' action may read
2528 * from. This set is determined by the clock vector's "happens before"
2530 * @param curr is the current ModelAction that we are exploring; it must be a
2533 void ModelChecker::build_reads_from_past(ModelAction *curr)
2535 std::vector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
2537 ASSERT(curr->is_read());
2539 ModelAction *last_sc_write = NULL;
2541 if (curr->is_seqcst())
2542 last_sc_write = get_last_seq_cst_write(curr);
2544 /* Iterate over all threads */
2545 for (i = 0; i < thrd_lists->size(); i++) {
2546 /* Iterate over actions in thread, starting from most recent */
2547 action_list_t *list = &(*thrd_lists)[i];
2548 action_list_t::reverse_iterator rit;
2549 for (rit = list->rbegin(); rit != list->rend(); rit++) {
2550 ModelAction *act = *rit;
2552 /* Only consider 'write' actions */
2553 if (!act->is_write() || act == curr)
2556 /* Don't consider more than one seq_cst write if we are a seq_cst read. */
2557 bool allow_read = true;
2559 if (curr->is_seqcst() && (act->is_seqcst() || (last_sc_write != NULL && act->happens_before(last_sc_write))) && act != last_sc_write)
2561 else if (curr->get_sleep_flag() && !curr->is_seqcst() && !sleep_can_read_from(curr, act))
2565 curr->get_node()->add_read_from(act);
2567 /* Include at most one act per-thread that "happens before" curr */
2568 if (act->happens_before(curr))
2573 if (DBG_ENABLED()) {
2574 model_print("Reached read action:\n");
2576 model_print("Printing may_read_from\n");
2577 curr->get_node()->print_may_read_from();
2578 model_print("End printing may_read_from\n");
2582 bool ModelChecker::sleep_can_read_from(ModelAction *curr, const ModelAction *write)
2585 /* UNINIT actions don't have a Node, and they never sleep */
2586 if (write->is_uninitialized())
2588 Node *prevnode = write->get_node()->get_parent();
2590 bool thread_sleep = prevnode->enabled_status(curr->get_tid()) == THREAD_SLEEP_SET;
2591 if (write->is_release() && thread_sleep)
2593 if (!write->is_rmw()) {
2596 if (write->get_reads_from() == NULL)
2598 write = write->get_reads_from();
2603 * @brief Create a new action representing an uninitialized atomic
2604 * @param location The memory location of the atomic object
2605 * @return A pointer to a new ModelAction
2607 ModelAction * ModelChecker::new_uninitialized_action(void *location) const
2609 ModelAction *act = (ModelAction *)snapshot_malloc(sizeof(class ModelAction));
2610 act = new (act) ModelAction(ATOMIC_UNINIT, std::memory_order_relaxed, location, 0, model_thread);
2611 act->create_cv(NULL);
2615 static void print_list(action_list_t *list)
2617 action_list_t::iterator it;
2619 model_print("---------------------------------------------------------------------\n");
2621 unsigned int hash = 0;
2623 for (it = list->begin(); it != list->end(); it++) {
2625 hash = hash^(hash<<3)^((*it)->hash());
2627 model_print("HASH %u\n", hash);
2628 model_print("---------------------------------------------------------------------\n");
2631 #if SUPPORT_MOD_ORDER_DUMP
2632 void ModelChecker::dumpGraph(char *filename) const
2635 sprintf(buffer, "%s.dot", filename);
2636 FILE *file = fopen(buffer, "w");
2637 fprintf(file, "digraph %s {\n", filename);
2638 mo_graph->dumpNodes(file);
2639 ModelAction **thread_array = (ModelAction **)model_calloc(1, sizeof(ModelAction *) * get_num_threads());
2641 for (action_list_t::iterator it = action_trace->begin(); it != action_trace->end(); it++) {
2642 ModelAction *action = *it;
2643 if (action->is_read()) {
2644 fprintf(file, "N%u [label=\"%u, T%u\"];\n", action->get_seq_number(), action->get_seq_number(), action->get_tid());
2645 if (action->get_reads_from() != NULL)
2646 fprintf(file, "N%u -> N%u[label=\"rf\", color=red];\n", action->get_seq_number(), action->get_reads_from()->get_seq_number());
2648 if (thread_array[action->get_tid()] != NULL) {
2649 fprintf(file, "N%u -> N%u[label=\"sb\", color=blue];\n", thread_array[action->get_tid()]->get_seq_number(), action->get_seq_number());
2652 thread_array[action->get_tid()] = action;
2654 fprintf(file, "}\n");
2655 model_free(thread_array);
2660 /** @brief Prints an execution trace summary. */
2661 void ModelChecker::print_summary() const
2663 #if SUPPORT_MOD_ORDER_DUMP
2665 char buffername[100];
2666 sprintf(buffername, "exec%04u", stats.num_total);
2667 mo_graph->dumpGraphToFile(buffername);
2668 sprintf(buffername, "graph%04u", stats.num_total);
2669 dumpGraph(buffername);
2672 model_print("Execution %d:", stats.num_total);
2673 if (isfeasibleprefix())
2676 print_infeasibility(" INFEASIBLE");
2677 print_list(action_trace);
2682 * Add a Thread to the system for the first time. Should only be called once
2684 * @param t The Thread to add
2686 void ModelChecker::add_thread(Thread *t)
2688 thread_map->put(id_to_int(t->get_id()), t);
2689 scheduler->add_thread(t);
2693 * Removes a thread from the scheduler.
2694 * @param the thread to remove.
2696 void ModelChecker::remove_thread(Thread *t)
2698 scheduler->remove_thread(t);
2702 * @brief Get a Thread reference by its ID
2703 * @param tid The Thread's ID
2704 * @return A Thread reference
2706 Thread * ModelChecker::get_thread(thread_id_t tid) const
2708 return thread_map->get(id_to_int(tid));
2712 * @brief Get a reference to the Thread in which a ModelAction was executed
2713 * @param act The ModelAction
2714 * @return A Thread reference
2716 Thread * ModelChecker::get_thread(const ModelAction *act) const
2718 return get_thread(act->get_tid());
2722 * @brief Check if a Thread is currently enabled
2723 * @param t The Thread to check
2724 * @return True if the Thread is currently enabled
2726 bool ModelChecker::is_enabled(Thread *t) const
2728 return scheduler->is_enabled(t);
2732 * @brief Check if a Thread is currently enabled
2733 * @param tid The ID of the Thread to check
2734 * @return True if the Thread is currently enabled
2736 bool ModelChecker::is_enabled(thread_id_t tid) const
2738 return scheduler->is_enabled(tid);
2742 * Switch from a user-context to the "master thread" context (a.k.a. system
2743 * context). This switch is made with the intention of exploring a particular
2744 * model-checking action (described by a ModelAction object). Must be called
2745 * from a user-thread context.
2747 * @param act The current action that will be explored. May be NULL only if
2748 * trace is exiting via an assertion (see ModelChecker::set_assert and
2749 * ModelChecker::has_asserted).
2750 * @return Return the value returned by the current action
2752 uint64_t ModelChecker::switch_to_master(ModelAction *act)
2755 Thread *old = thread_current();
2756 set_current_action(act);
2757 old->set_state(THREAD_READY);
2758 if (Thread::swap(old, &system_context) < 0) {
2759 perror("swap threads");
2762 return old->get_return_value();
2766 * Takes the next step in the execution, if possible.
2767 * @param curr The current step to take
2768 * @return Returns true (success) if a step was taken and false otherwise.
2770 bool ModelChecker::take_step(ModelAction *curr)
2775 Thread *curr_thrd = get_thread(curr);
2776 ASSERT(curr_thrd->get_state() == THREAD_READY);
2778 curr = check_current_action(curr);
2780 /* Infeasible -> don't take any more steps */
2781 if (is_infeasible())
2783 else if (isfeasibleprefix() && have_bug_reports()) {
2788 if (params.bound != 0)
2789 if (priv->used_sequence_numbers > params.bound)
2792 if (curr_thrd->is_blocked() || curr_thrd->is_complete())
2793 scheduler->remove_thread(curr_thrd);
2795 Thread *next_thrd = get_next_thread(curr);
2796 next_thrd = scheduler->next_thread(next_thrd);
2798 DEBUG("(%d, %d)\n", curr_thrd ? id_to_int(curr_thrd->get_id()) : -1,
2799 next_thrd ? id_to_int(next_thrd->get_id()) : -1);
2802 * Launch end-of-execution release sequence fixups only when there are:
2804 * (1) no more user threads to run (or when execution replay chooses
2805 * the 'model_thread')
2806 * (2) pending release sequences
2807 * (3) pending assertions (i.e., data races)
2808 * (4) no pending promises
2810 if (!pending_rel_seqs->empty() && (!next_thrd || next_thrd->is_model_thread()) &&
2811 is_feasible_prefix_ignore_relseq() && !unrealizedraces.empty()) {
2812 model_print("*** WARNING: release sequence fixup action (%zu pending release seuqences) ***\n",
2813 pending_rel_seqs->size());
2814 ModelAction *fixup = new ModelAction(MODEL_FIXUP_RELSEQ,
2815 std::memory_order_seq_cst, NULL, VALUE_NONE,
2817 set_current_action(fixup);
2821 /* next_thrd == NULL -> don't take any more steps */
2825 next_thrd->set_state(THREAD_RUNNING);
2827 if (next_thrd->get_pending() != NULL) {
2828 /* restart a pending action */
2829 set_current_action(next_thrd->get_pending());
2830 next_thrd->set_pending(NULL);
2831 next_thrd->set_state(THREAD_READY);
2835 /* Return false only if swap fails with an error */
2836 return (Thread::swap(&system_context, next_thrd) == 0);
2839 /** Wrapper to run the user's main function, with appropriate arguments */
2840 void user_main_wrapper(void *)
2842 user_main(model->params.argc, model->params.argv);
2845 /** @brief Run ModelChecker for the user program */
2846 void ModelChecker::run()
2850 Thread *t = new Thread(&user_thread, &user_main_wrapper, NULL);
2854 /* Run user thread up to its first action */
2855 scheduler->next_thread(t);
2856 Thread::swap(&system_context, t);
2858 /* Wait for all threads to complete */
2859 while (take_step(priv->current_action));
2860 } while (next_execution());