11 #include "snapshot-interface.h"
13 #include "clockvector.h"
14 #include "cyclegraph.h"
17 #include "threads-model.h"
19 #include "traceanalysis.h"
20 #include "bugmessage.h"
22 #define INITIAL_THREAD_ID 0
27 * Structure for holding small ModelChecker members that should be snapshotted
29 struct model_snapshot_members {
30 model_snapshot_members() :
31 /* First thread created will have id INITIAL_THREAD_ID */
32 next_thread_id(INITIAL_THREAD_ID),
33 used_sequence_numbers(0),
37 failed_promise(false),
38 too_many_reads(false),
39 no_valid_reads(false),
40 bad_synchronization(false),
44 ~model_snapshot_members() {
45 for (unsigned int i = 0; i < bugs.size(); i++)
50 unsigned int next_thread_id;
51 modelclock_t used_sequence_numbers;
52 ModelAction *next_backtrack;
53 SnapVector<bug_message *> bugs;
54 struct execution_stats stats;
58 /** @brief Incorrectly-ordered synchronization was made */
59 bool bad_synchronization;
65 /** @brief Constructor */
66 ModelChecker::ModelChecker(struct model_params params) :
67 /* Initialize default scheduler */
69 scheduler(new Scheduler()),
71 earliest_diverge(NULL),
72 action_trace(new action_list_t()),
73 thread_map(new HashTable<int, Thread *, int>()),
74 obj_map(new HashTable<const void *, action_list_t *, uintptr_t, 4>()),
75 condvar_waiters_map(new HashTable<const void *, action_list_t *, uintptr_t, 4>()),
76 obj_thrd_map(new HashTable<void *, SnapVector<action_list_t> *, uintptr_t, 4 >()),
77 promises(new SnapVector<Promise *>()),
78 futurevalues(new SnapVector<struct PendingFutureValue>()),
79 pending_rel_seqs(new SnapVector<struct release_seq *>()),
80 thrd_last_action(new SnapVector<ModelAction *>(1)),
81 thrd_last_fence_release(new SnapVector<ModelAction *>()),
82 node_stack(new NodeStack()),
83 trace_analyses(new ModelVector<Trace_Analysis *>()),
84 priv(new struct model_snapshot_members()),
85 mo_graph(new CycleGraph())
87 /* Initialize a model-checker thread, for special ModelActions */
88 model_thread = new Thread(get_next_id());
89 thread_map->put(id_to_int(model_thread->get_id()), model_thread);
92 /** @brief Destructor */
93 ModelChecker::~ModelChecker()
95 for (unsigned int i = 0; i < get_num_threads(); i++)
96 delete thread_map->get(i);
101 delete condvar_waiters_map;
104 for (unsigned int i = 0; i < promises->size(); i++)
105 delete (*promises)[i];
108 delete pending_rel_seqs;
110 delete thrd_last_action;
111 delete thrd_last_fence_release;
113 for (unsigned int i = 0; i < trace_analyses->size(); i++)
114 delete (*trace_analyses)[i];
115 delete trace_analyses;
121 static action_list_t * get_safe_ptr_action(HashTable<const void *, action_list_t *, uintptr_t, 4> * hash, void * ptr)
123 action_list_t *tmp = hash->get(ptr);
125 tmp = new action_list_t();
131 static SnapVector<action_list_t> * get_safe_ptr_vect_action(HashTable<void *, SnapVector<action_list_t> *, uintptr_t, 4> * hash, void * ptr)
133 SnapVector<action_list_t> *tmp = hash->get(ptr);
135 tmp = new SnapVector<action_list_t>();
141 action_list_t * ModelChecker::get_actions_on_obj(void * obj, thread_id_t tid) {
142 SnapVector<action_list_t> *wrv=obj_thrd_map->get(obj);
145 unsigned int thread=id_to_int(tid);
146 if (thread < wrv->size())
147 return &(*wrv)[thread];
154 * Restores user program to initial state and resets all model-checker data
157 void ModelChecker::reset_to_initial_state()
159 DEBUG("+++ Resetting to initial state +++\n");
160 node_stack->reset_execution();
163 * FIXME: if we utilize partial rollback, we will need to free only
164 * those pending actions which were NOT pending before the rollback
167 for (unsigned int i = 0; i < get_num_threads(); i++)
168 delete get_thread(int_to_id(i))->get_pending();
170 snapshot_backtrack_before(0);
173 /** @return a thread ID for a new Thread */
174 thread_id_t ModelChecker::get_next_id()
176 return priv->next_thread_id++;
179 /** @return the number of user threads created during this execution */
180 unsigned int ModelChecker::get_num_threads() const
182 return priv->next_thread_id;
186 * Must be called from user-thread context (e.g., through the global
187 * thread_current() interface)
189 * @return The currently executing Thread.
191 Thread * ModelChecker::get_current_thread() const
193 return scheduler->get_current_thread();
196 /** @return a sequence number for a new ModelAction */
197 modelclock_t ModelChecker::get_next_seq_num()
199 return ++priv->used_sequence_numbers;
202 Node * ModelChecker::get_curr_node() const
204 return node_stack->get_head();
208 * @brief Select the next thread to execute based on the curren action
210 * RMW actions occur in two parts, and we cannot split them. And THREAD_CREATE
211 * actions should be followed by the execution of their child thread. In either
212 * case, the current action should determine the next thread schedule.
214 * @param curr The current action
215 * @return The next thread to run, if the current action will determine this
216 * selection; otherwise NULL
218 Thread * ModelChecker::action_select_next_thread(const ModelAction *curr) const
220 /* Do not split atomic RMW */
222 return get_thread(curr);
223 /* Follow CREATE with the created thread */
224 if (curr->get_type() == THREAD_CREATE)
225 return curr->get_thread_operand();
230 * @brief Choose the next thread to execute.
232 * This function chooses the next thread that should execute. It can enforce
233 * execution replay/backtracking or, if the model-checker has no preference
234 * regarding the next thread (i.e., when exploring a new execution ordering),
235 * we defer to the scheduler.
237 * @return The next chosen thread to run, if any exist. Or else if the current
238 * execution should terminate, return NULL.
240 Thread * ModelChecker::get_next_thread()
245 * Have we completed exploring the preselected path? Then let the
249 return scheduler->select_next_thread();
251 /* Else, we are trying to replay an execution */
252 ModelAction *next = node_stack->get_next()->get_action();
254 if (next == diverge) {
255 if (earliest_diverge == NULL || *diverge < *earliest_diverge)
256 earliest_diverge = diverge;
258 Node *nextnode = next->get_node();
259 Node *prevnode = nextnode->get_parent();
260 scheduler->update_sleep_set(prevnode);
262 /* Reached divergence point */
263 if (nextnode->increment_behaviors()) {
264 /* Execute the same thread with a new behavior */
265 tid = next->get_tid();
266 node_stack->pop_restofstack(2);
269 /* Make a different thread execute for next step */
270 scheduler->add_sleep(get_thread(next->get_tid()));
271 tid = prevnode->get_next_backtrack();
272 /* Make sure the backtracked thread isn't sleeping. */
273 node_stack->pop_restofstack(1);
274 if (diverge == earliest_diverge) {
275 earliest_diverge = prevnode->get_action();
278 /* Start the round robin scheduler from this thread id */
279 scheduler->set_scheduler_thread(tid);
280 /* The correct sleep set is in the parent node. */
283 DEBUG("*** Divergence point ***\n");
287 tid = next->get_tid();
289 DEBUG("*** ModelChecker chose next thread = %d ***\n", id_to_int(tid));
290 ASSERT(tid != THREAD_ID_T_NONE);
291 return get_thread(id_to_int(tid));
295 * We need to know what the next actions of all threads in the sleep
296 * set will be. This method computes them and stores the actions at
297 * the corresponding thread object's pending action.
300 void ModelChecker::execute_sleep_set()
302 for (unsigned int i = 0; i < get_num_threads(); i++) {
303 thread_id_t tid = int_to_id(i);
304 Thread *thr = get_thread(tid);
305 if (scheduler->is_sleep_set(thr) && thr->get_pending()) {
306 thr->get_pending()->set_sleep_flag();
312 * @brief Should the current action wake up a given thread?
314 * @param curr The current action
315 * @param thread The thread that we might wake up
316 * @return True, if we should wake up the sleeping thread; false otherwise
318 bool ModelChecker::should_wake_up(const ModelAction *curr, const Thread *thread) const
320 const ModelAction *asleep = thread->get_pending();
321 /* Don't allow partial RMW to wake anyone up */
324 /* Synchronizing actions may have been backtracked */
325 if (asleep->could_synchronize_with(curr))
327 /* All acquire/release fences and fence-acquire/store-release */
328 if (asleep->is_fence() && asleep->is_acquire() && curr->is_release())
330 /* Fence-release + store can awake load-acquire on the same location */
331 if (asleep->is_read() && asleep->is_acquire() && curr->same_var(asleep) && curr->is_write()) {
332 ModelAction *fence_release = get_last_fence_release(curr->get_tid());
333 if (fence_release && *(get_last_action(thread->get_id())) < *fence_release)
339 void ModelChecker::wake_up_sleeping_actions(ModelAction *curr)
341 for (unsigned int i = 0; i < get_num_threads(); i++) {
342 Thread *thr = get_thread(int_to_id(i));
343 if (scheduler->is_sleep_set(thr)) {
344 if (should_wake_up(curr, thr))
345 /* Remove this thread from sleep set */
346 scheduler->remove_sleep(thr);
351 /** @brief Alert the model-checker that an incorrectly-ordered
352 * synchronization was made */
353 void ModelChecker::set_bad_synchronization()
355 priv->bad_synchronization = true;
359 * Check whether the current trace has triggered an assertion which should halt
362 * @return True, if the execution should be aborted; false otherwise
364 bool ModelChecker::has_asserted() const
366 return priv->asserted;
370 * Trigger a trace assertion which should cause this execution to be halted.
371 * This can be due to a detected bug or due to an infeasibility that should
374 void ModelChecker::set_assert()
376 priv->asserted = true;
380 * Check if we are in a deadlock. Should only be called at the end of an
381 * execution, although it should not give false positives in the middle of an
382 * execution (there should be some ENABLED thread).
384 * @return True if program is in a deadlock; false otherwise
386 bool ModelChecker::is_deadlocked() const
388 bool blocking_threads = false;
389 for (unsigned int i = 0; i < get_num_threads(); i++) {
390 thread_id_t tid = int_to_id(i);
393 Thread *t = get_thread(tid);
394 if (!t->is_model_thread() && t->get_pending())
395 blocking_threads = true;
397 return blocking_threads;
401 * Check if this is a complete execution. That is, have all thread completed
402 * execution (rather than exiting because sleep sets have forced a redundant
405 * @return True if the execution is complete.
407 bool ModelChecker::is_complete_execution() const
409 for (unsigned int i = 0; i < get_num_threads(); i++)
410 if (is_enabled(int_to_id(i)))
416 * @brief Assert a bug in the executing program.
418 * Use this function to assert any sort of bug in the user program. If the
419 * current trace is feasible (actually, a prefix of some feasible execution),
420 * then this execution will be aborted, printing the appropriate message. If
421 * the current trace is not yet feasible, the error message will be stashed and
422 * printed if the execution ever becomes feasible.
424 * @param msg Descriptive message for the bug (do not include newline char)
425 * @return True if bug is immediately-feasible
427 bool ModelChecker::assert_bug(const char *msg, ...)
433 vsnprintf(str, sizeof(str), msg, ap);
436 priv->bugs.push_back(new bug_message(str));
438 if (isfeasibleprefix()) {
446 * @brief Assert a bug in the executing program, asserted by a user thread
447 * @see ModelChecker::assert_bug
448 * @param msg Descriptive message for the bug (do not include newline char)
450 void ModelChecker::assert_user_bug(const char *msg)
452 /* If feasible bug, bail out now */
454 switch_to_master(NULL);
457 /** @return True, if any bugs have been reported for this execution */
458 bool ModelChecker::have_bug_reports() const
460 return priv->bugs.size() != 0;
463 /** @brief Print bug report listing for this execution (if any bugs exist) */
464 void ModelChecker::print_bugs() const
466 if (have_bug_reports()) {
467 model_print("Bug report: %zu bug%s detected\n",
469 priv->bugs.size() > 1 ? "s" : "");
470 for (unsigned int i = 0; i < priv->bugs.size(); i++)
471 priv->bugs[i]->print();
476 * @brief Record end-of-execution stats
478 * Must be run when exiting an execution. Records various stats.
479 * @see struct execution_stats
481 void ModelChecker::record_stats()
484 if (!isfeasibleprefix())
485 stats.num_infeasible++;
486 else if (have_bug_reports())
487 stats.num_buggy_executions++;
488 else if (is_complete_execution())
489 stats.num_complete++;
491 stats.num_redundant++;
494 * @todo We can violate this ASSERT() when fairness/sleep sets
495 * conflict to cause an execution to terminate, e.g. with:
496 * Scheduler: [0: disabled][1: disabled][2: sleep][3: current, enabled]
498 //ASSERT(scheduler->all_threads_sleeping());
502 /** @brief Print execution stats */
503 void ModelChecker::print_stats() const
505 model_print("Number of complete, bug-free executions: %d\n", stats.num_complete);
506 model_print("Number of redundant executions: %d\n", stats.num_redundant);
507 model_print("Number of buggy executions: %d\n", stats.num_buggy_executions);
508 model_print("Number of infeasible executions: %d\n", stats.num_infeasible);
509 model_print("Total executions: %d\n", stats.num_total);
510 model_print("Total nodes created: %d\n", node_stack->get_total_nodes());
514 * @brief End-of-exeuction print
515 * @param printbugs Should any existing bugs be printed?
517 void ModelChecker::print_execution(bool printbugs) const
519 print_program_output();
521 if (params.verbose) {
522 model_print("Earliest divergence point since last feasible execution:\n");
523 if (earliest_diverge)
524 earliest_diverge->print();
526 model_print("(Not set)\n");
532 /* Don't print invalid bugs */
541 * Queries the model-checker for more executions to explore and, if one
542 * exists, resets the model-checker state to execute a new execution.
544 * @return If there are more executions to explore, return true. Otherwise,
547 bool ModelChecker::next_execution()
550 /* Is this execution a feasible execution that's worth bug-checking? */
551 bool complete = isfeasibleprefix() && (is_complete_execution() ||
554 /* End-of-execution bug checks */
557 assert_bug("Deadlock detected");
560 run_trace_analyses();
566 if (params.verbose || (complete && have_bug_reports()))
567 print_execution(complete);
569 clear_program_output();
572 earliest_diverge = NULL;
574 if ((diverge = get_next_backtrack()) == NULL)
578 model_print("Next execution will diverge at:\n");
582 reset_to_initial_state();
586 /** @brief Run trace analyses on complete trace */
587 void ModelChecker::run_trace_analyses() {
588 for (unsigned int i = 0; i < trace_analyses->size(); i++)
589 (*trace_analyses)[i]->analyze(action_trace);
593 * @brief Find the last fence-related backtracking conflict for a ModelAction
595 * This function performs the search for the most recent conflicting action
596 * against which we should perform backtracking, as affected by fence
597 * operations. This includes pairs of potentially-synchronizing actions which
598 * occur due to fence-acquire or fence-release, and hence should be explored in
599 * the opposite execution order.
601 * @param act The current action
602 * @return The most recent action which conflicts with act due to fences
604 ModelAction * ModelChecker::get_last_fence_conflict(ModelAction *act) const
606 /* Only perform release/acquire fence backtracking for stores */
607 if (!act->is_write())
610 /* Find a fence-release (or, act is a release) */
611 ModelAction *last_release;
612 if (act->is_release())
615 last_release = get_last_fence_release(act->get_tid());
619 /* Skip past the release */
620 action_list_t *list = action_trace;
621 action_list_t::reverse_iterator rit;
622 for (rit = list->rbegin(); rit != list->rend(); rit++)
623 if (*rit == last_release)
625 ASSERT(rit != list->rend());
630 * load --sb-> fence-acquire */
631 ModelVector<ModelAction *> acquire_fences(get_num_threads(), NULL);
632 ModelVector<ModelAction *> prior_loads(get_num_threads(), NULL);
633 bool found_acquire_fences = false;
634 for ( ; rit != list->rend(); rit++) {
635 ModelAction *prev = *rit;
636 if (act->same_thread(prev))
639 int tid = id_to_int(prev->get_tid());
641 if (prev->is_read() && act->same_var(prev)) {
642 if (prev->is_acquire()) {
643 /* Found most recent load-acquire, don't need
644 * to search for more fences */
645 if (!found_acquire_fences)
648 prior_loads[tid] = prev;
651 if (prev->is_acquire() && prev->is_fence() && !acquire_fences[tid]) {
652 found_acquire_fences = true;
653 acquire_fences[tid] = prev;
657 ModelAction *latest_backtrack = NULL;
658 for (unsigned int i = 0; i < acquire_fences.size(); i++)
659 if (acquire_fences[i] && prior_loads[i])
660 if (!latest_backtrack || *latest_backtrack < *acquire_fences[i])
661 latest_backtrack = acquire_fences[i];
662 return latest_backtrack;
666 * @brief Find the last backtracking conflict for a ModelAction
668 * This function performs the search for the most recent conflicting action
669 * against which we should perform backtracking. This primary includes pairs of
670 * synchronizing actions which should be explored in the opposite execution
673 * @param act The current action
674 * @return The most recent action which conflicts with act
676 ModelAction * ModelChecker::get_last_conflict(ModelAction *act) const
678 switch (act->get_type()) {
679 /* case ATOMIC_FENCE: fences don't directly cause backtracking */
683 ModelAction *ret = NULL;
685 /* linear search: from most recent to oldest */
686 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
687 action_list_t::reverse_iterator rit;
688 for (rit = list->rbegin(); rit != list->rend(); rit++) {
689 ModelAction *prev = *rit;
690 if (prev->could_synchronize_with(act)) {
696 ModelAction *ret2 = get_last_fence_conflict(act);
706 case ATOMIC_TRYLOCK: {
707 /* linear search: from most recent to oldest */
708 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
709 action_list_t::reverse_iterator rit;
710 for (rit = list->rbegin(); rit != list->rend(); rit++) {
711 ModelAction *prev = *rit;
712 if (act->is_conflicting_lock(prev))
717 case ATOMIC_UNLOCK: {
718 /* linear search: from most recent to oldest */
719 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
720 action_list_t::reverse_iterator rit;
721 for (rit = list->rbegin(); rit != list->rend(); rit++) {
722 ModelAction *prev = *rit;
723 if (!act->same_thread(prev) && prev->is_failed_trylock())
729 /* linear search: from most recent to oldest */
730 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
731 action_list_t::reverse_iterator rit;
732 for (rit = list->rbegin(); rit != list->rend(); rit++) {
733 ModelAction *prev = *rit;
734 if (!act->same_thread(prev) && prev->is_failed_trylock())
736 if (!act->same_thread(prev) && prev->is_notify())
742 case ATOMIC_NOTIFY_ALL:
743 case ATOMIC_NOTIFY_ONE: {
744 /* linear search: from most recent to oldest */
745 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
746 action_list_t::reverse_iterator rit;
747 for (rit = list->rbegin(); rit != list->rend(); rit++) {
748 ModelAction *prev = *rit;
749 if (!act->same_thread(prev) && prev->is_wait())
760 /** This method finds backtracking points where we should try to
761 * reorder the parameter ModelAction against.
763 * @param the ModelAction to find backtracking points for.
765 void ModelChecker::set_backtracking(ModelAction *act)
767 Thread *t = get_thread(act);
768 ModelAction *prev = get_last_conflict(act);
772 Node *node = prev->get_node()->get_parent();
774 /* See Dynamic Partial Order Reduction (addendum), POPL '05 */
775 int low_tid, high_tid;
776 if (node->enabled_status(t->get_id()) == THREAD_ENABLED) {
777 low_tid = id_to_int(act->get_tid());
778 high_tid = low_tid + 1;
781 high_tid = get_num_threads();
784 for (int i = low_tid; i < high_tid; i++) {
785 thread_id_t tid = int_to_id(i);
787 /* Make sure this thread can be enabled here. */
788 if (i >= node->get_num_threads())
791 /* See Dynamic Partial Order Reduction (addendum), POPL '05 */
792 /* Don't backtrack into a point where the thread is disabled or sleeping. */
793 if (node->enabled_status(tid) != THREAD_ENABLED)
796 /* Check if this has been explored already */
797 if (node->has_been_explored(tid))
800 /* See if fairness allows */
801 if (model->params.fairwindow != 0 && !node->has_priority(tid)) {
803 for (int t = 0; t < node->get_num_threads(); t++) {
804 thread_id_t tother = int_to_id(t);
805 if (node->is_enabled(tother) && node->has_priority(tother)) {
814 /* See if CHESS-like yield fairness allows */
815 if (model->params.yieldon) {
817 for (int t = 0; t < node->get_num_threads(); t++) {
818 thread_id_t tother = int_to_id(t);
819 if (node->is_enabled(tother) && node->has_priority_over(tid, tother)) {
828 /* Cache the latest backtracking point */
829 set_latest_backtrack(prev);
831 /* If this is a new backtracking point, mark the tree */
832 if (!node->set_backtrack(tid))
834 DEBUG("Setting backtrack: conflict = %d, instead tid = %d\n",
835 id_to_int(prev->get_tid()),
836 id_to_int(t->get_id()));
845 * @brief Cache the a backtracking point as the "most recent", if eligible
847 * Note that this does not prepare the NodeStack for this backtracking
848 * operation, it only caches the action on a per-execution basis
850 * @param act The operation at which we should explore a different next action
851 * (i.e., backtracking point)
852 * @return True, if this action is now the most recent backtracking point;
855 bool ModelChecker::set_latest_backtrack(ModelAction *act)
857 if (!priv->next_backtrack || *act > *priv->next_backtrack) {
858 priv->next_backtrack = act;
865 * Returns last backtracking point. The model checker will explore a different
866 * path for this point in the next execution.
867 * @return The ModelAction at which the next execution should diverge.
869 ModelAction * ModelChecker::get_next_backtrack()
871 ModelAction *next = priv->next_backtrack;
872 priv->next_backtrack = NULL;
877 * Processes a read model action.
878 * @param curr is the read model action to process.
879 * @return True if processing this read updates the mo_graph.
881 bool ModelChecker::process_read(ModelAction *curr)
883 Node *node = curr->get_node();
885 bool updated = false;
886 switch (node->get_read_from_status()) {
887 case READ_FROM_PAST: {
888 const ModelAction *rf = node->get_read_from_past();
891 mo_graph->startChanges();
893 ASSERT(!is_infeasible());
894 if (!check_recency(curr, rf)) {
895 if (node->increment_read_from()) {
896 mo_graph->rollbackChanges();
899 priv->too_many_reads = true;
903 updated = r_modification_order(curr, rf);
905 mo_graph->commitChanges();
906 mo_check_promises(curr, true);
909 case READ_FROM_PROMISE: {
910 Promise *promise = curr->get_node()->get_read_from_promise();
911 if (promise->add_reader(curr))
912 priv->failed_promise = true;
913 curr->set_read_from_promise(promise);
914 mo_graph->startChanges();
915 if (!check_recency(curr, promise))
916 priv->too_many_reads = true;
917 updated = r_modification_order(curr, promise);
918 mo_graph->commitChanges();
921 case READ_FROM_FUTURE: {
922 /* Read from future value */
923 struct future_value fv = node->get_future_value();
924 Promise *promise = new Promise(curr, fv);
925 curr->set_read_from_promise(promise);
926 promises->push_back(promise);
927 mo_graph->startChanges();
928 updated = r_modification_order(curr, promise);
929 mo_graph->commitChanges();
935 get_thread(curr)->set_return_value(curr->get_return_value());
941 * Processes a lock, trylock, or unlock model action. @param curr is
942 * the read model action to process.
944 * The try lock operation checks whether the lock is taken. If not,
945 * it falls to the normal lock operation case. If so, it returns
948 * The lock operation has already been checked that it is enabled, so
949 * it just grabs the lock and synchronizes with the previous unlock.
951 * The unlock operation has to re-enable all of the threads that are
952 * waiting on the lock.
954 * @return True if synchronization was updated; false otherwise
956 bool ModelChecker::process_mutex(ModelAction *curr)
958 std::mutex *mutex = curr->get_mutex();
959 struct std::mutex_state *state = NULL;
962 state = mutex->get_state();
964 switch (curr->get_type()) {
965 case ATOMIC_TRYLOCK: {
966 bool success = !state->locked;
967 curr->set_try_lock(success);
969 get_thread(curr)->set_return_value(0);
972 get_thread(curr)->set_return_value(1);
974 //otherwise fall into the lock case
976 if (curr->get_cv()->getClock(state->alloc_tid) <= state->alloc_clock)
977 assert_bug("Lock access before initialization");
978 state->locked = get_thread(curr);
979 ModelAction *unlock = get_last_unlock(curr);
980 //synchronize with the previous unlock statement
981 if (unlock != NULL) {
982 curr->synchronize_with(unlock);
988 case ATOMIC_UNLOCK: {
989 /* wake up the other threads */
990 for (unsigned int i = 0; i < get_num_threads(); i++) {
991 Thread *t = get_thread(int_to_id(i));
992 Thread *curr_thrd = get_thread(curr);
993 if (t->waiting_on() == curr_thrd && t->get_pending()->is_lock())
997 /* unlock the lock - after checking who was waiting on it */
998 state->locked = NULL;
1000 if (!curr->is_wait())
1001 break; /* The rest is only for ATOMIC_WAIT */
1003 /* Should we go to sleep? (simulate spurious failures) */
1004 if (curr->get_node()->get_misc() == 0) {
1005 get_safe_ptr_action(condvar_waiters_map, curr->get_location())->push_back(curr);
1007 scheduler->sleep(get_thread(curr));
1011 case ATOMIC_NOTIFY_ALL: {
1012 action_list_t *waiters = get_safe_ptr_action(condvar_waiters_map, curr->get_location());
1013 //activate all the waiting threads
1014 for (action_list_t::iterator rit = waiters->begin(); rit != waiters->end(); rit++) {
1015 scheduler->wake(get_thread(*rit));
1020 case ATOMIC_NOTIFY_ONE: {
1021 action_list_t *waiters = get_safe_ptr_action(condvar_waiters_map, curr->get_location());
1022 int wakeupthread = curr->get_node()->get_misc();
1023 action_list_t::iterator it = waiters->begin();
1024 advance(it, wakeupthread);
1025 scheduler->wake(get_thread(*it));
1037 * @brief Check if the current pending promises allow a future value to be sent
1039 * If one of the following is true:
1040 * (a) there are no pending promises
1041 * (b) the reader and writer do not cross any promises
1042 * Then, it is safe to pass a future value back now.
1044 * Otherwise, we must save the pending future value until (a) or (b) is true
1046 * @param writer The operation which sends the future value. Must be a write.
1047 * @param reader The operation which will observe the value. Must be a read.
1048 * @return True if the future value can be sent now; false if it must wait.
1050 bool ModelChecker::promises_may_allow(const ModelAction *writer,
1051 const ModelAction *reader) const
1053 if (promises->empty())
1055 for(int i=promises->size()-1;i>=0;i--) {
1056 ModelAction *pr=(*promises)[i]->get_reader(0);
1057 //reader is after promise...doesn't cross any promise
1060 //writer is after promise, reader before...bad...
1068 * @brief Add a future value to a reader
1070 * This function performs a few additional checks to ensure that the future
1071 * value can be feasibly observed by the reader
1073 * @param writer The operation whose value is sent. Must be a write.
1074 * @param reader The read operation which may read the future value. Must be a read.
1076 void ModelChecker::add_future_value(const ModelAction *writer, ModelAction *reader)
1078 /* Do more ambitious checks now that mo is more complete */
1079 if (!mo_may_allow(writer, reader))
1082 Node *node = reader->get_node();
1084 /* Find an ancestor thread which exists at the time of the reader */
1085 Thread *write_thread = get_thread(writer);
1086 while (id_to_int(write_thread->get_id()) >= node->get_num_threads())
1087 write_thread = write_thread->get_parent();
1089 struct future_value fv = {
1090 writer->get_write_value(),
1091 writer->get_seq_number() + params.maxfuturedelay,
1092 write_thread->get_id(),
1094 if (node->add_future_value(fv))
1095 set_latest_backtrack(reader);
1099 * Process a write ModelAction
1100 * @param curr The ModelAction to process
1101 * @return True if the mo_graph was updated or promises were resolved
1103 bool ModelChecker::process_write(ModelAction *curr)
1105 /* Readers to which we may send our future value */
1106 ModelVector<ModelAction *> send_fv;
1108 const ModelAction *earliest_promise_reader;
1109 bool updated_promises = false;
1111 bool updated_mod_order = w_modification_order(curr, &send_fv);
1112 Promise *promise = pop_promise_to_resolve(curr);
1115 earliest_promise_reader = promise->get_reader(0);
1116 updated_promises = resolve_promise(curr, promise);
1118 earliest_promise_reader = NULL;
1120 for (unsigned int i = 0; i < send_fv.size(); i++) {
1121 ModelAction *read = send_fv[i];
1123 /* Don't send future values to reads after the Promise we resolve */
1124 if (!earliest_promise_reader || *read < *earliest_promise_reader) {
1125 /* Check if future value can be sent immediately */
1126 if (promises_may_allow(curr, read)) {
1127 add_future_value(curr, read);
1129 futurevalues->push_back(PendingFutureValue(curr, read));
1134 /* Check the pending future values */
1135 for (int i = (int)futurevalues->size() - 1; i >= 0; i--) {
1136 struct PendingFutureValue pfv = (*futurevalues)[i];
1137 if (promises_may_allow(pfv.writer, pfv.reader)) {
1138 add_future_value(pfv.writer, pfv.reader);
1139 futurevalues->erase(futurevalues->begin() + i);
1143 mo_graph->commitChanges();
1144 mo_check_promises(curr, false);
1146 get_thread(curr)->set_return_value(VALUE_NONE);
1147 return updated_mod_order || updated_promises;
1151 * Process a fence ModelAction
1152 * @param curr The ModelAction to process
1153 * @return True if synchronization was updated
1155 bool ModelChecker::process_fence(ModelAction *curr)
1158 * fence-relaxed: no-op
1159 * fence-release: only log the occurence (not in this function), for
1160 * use in later synchronization
1161 * fence-acquire (this function): search for hypothetical release
1163 * fence-seq-cst: MO constraints formed in {r,w}_modification_order
1165 bool updated = false;
1166 if (curr->is_acquire()) {
1167 action_list_t *list = action_trace;
1168 action_list_t::reverse_iterator rit;
1169 /* Find X : is_read(X) && X --sb-> curr */
1170 for (rit = list->rbegin(); rit != list->rend(); rit++) {
1171 ModelAction *act = *rit;
1174 if (act->get_tid() != curr->get_tid())
1176 /* Stop at the beginning of the thread */
1177 if (act->is_thread_start())
1179 /* Stop once we reach a prior fence-acquire */
1180 if (act->is_fence() && act->is_acquire())
1182 if (!act->is_read())
1184 /* read-acquire will find its own release sequences */
1185 if (act->is_acquire())
1188 /* Establish hypothetical release sequences */
1189 rel_heads_list_t release_heads;
1190 get_release_seq_heads(curr, act, &release_heads);
1191 for (unsigned int i = 0; i < release_heads.size(); i++)
1192 if (!curr->synchronize_with(release_heads[i]))
1193 set_bad_synchronization();
1194 if (release_heads.size() != 0)
1202 * @brief Process the current action for thread-related activity
1204 * Performs current-action processing for a THREAD_* ModelAction. Proccesses
1205 * may include setting Thread status, completing THREAD_FINISH/THREAD_JOIN
1206 * synchronization, etc. This function is a no-op for non-THREAD actions
1207 * (e.g., ATOMIC_{READ,WRITE,RMW,LOCK}, etc.)
1209 * @param curr The current action
1210 * @return True if synchronization was updated or a thread completed
1212 bool ModelChecker::process_thread_action(ModelAction *curr)
1214 bool updated = false;
1216 switch (curr->get_type()) {
1217 case THREAD_CREATE: {
1218 thrd_t *thrd = (thrd_t *)curr->get_location();
1219 struct thread_params *params = (struct thread_params *)curr->get_value();
1220 Thread *th = new Thread(thrd, params->func, params->arg, get_thread(curr));
1222 th->set_creation(curr);
1223 /* Promises can be satisfied by children */
1224 for (unsigned int i = 0; i < promises->size(); i++) {
1225 Promise *promise = (*promises)[i];
1226 if (promise->thread_is_available(curr->get_tid()))
1227 promise->add_thread(th->get_id());
1232 Thread *blocking = curr->get_thread_operand();
1233 ModelAction *act = get_last_action(blocking->get_id());
1234 curr->synchronize_with(act);
1235 updated = true; /* trigger rel-seq checks */
1238 case THREAD_FINISH: {
1239 Thread *th = get_thread(curr);
1240 /* Wake up any joining threads */
1241 for (unsigned int i = 0; i < get_num_threads(); i++) {
1242 Thread *waiting = get_thread(int_to_id(i));
1243 if (waiting->waiting_on() == th &&
1244 waiting->get_pending()->is_thread_join())
1245 scheduler->wake(waiting);
1248 /* Completed thread can't satisfy promises */
1249 for (unsigned int i = 0; i < promises->size(); i++) {
1250 Promise *promise = (*promises)[i];
1251 if (promise->thread_is_available(th->get_id()))
1252 if (promise->eliminate_thread(th->get_id()))
1253 priv->failed_promise = true;
1255 updated = true; /* trigger rel-seq checks */
1258 case THREAD_START: {
1259 check_promises(curr->get_tid(), NULL, curr->get_cv());
1270 * @brief Process the current action for release sequence fixup activity
1272 * Performs model-checker release sequence fixups for the current action,
1273 * forcing a single pending release sequence to break (with a given, potential
1274 * "loose" write) or to complete (i.e., synchronize). If a pending release
1275 * sequence forms a complete release sequence, then we must perform the fixup
1276 * synchronization, mo_graph additions, etc.
1278 * @param curr The current action; must be a release sequence fixup action
1279 * @param work_queue The work queue to which to add work items as they are
1282 void ModelChecker::process_relseq_fixup(ModelAction *curr, work_queue_t *work_queue)
1284 const ModelAction *write = curr->get_node()->get_relseq_break();
1285 struct release_seq *sequence = pending_rel_seqs->back();
1286 pending_rel_seqs->pop_back();
1288 ModelAction *acquire = sequence->acquire;
1289 const ModelAction *rf = sequence->rf;
1290 const ModelAction *release = sequence->release;
1294 ASSERT(release->same_thread(rf));
1296 if (write == NULL) {
1298 * @todo Forcing a synchronization requires that we set
1299 * modification order constraints. For instance, we can't allow
1300 * a fixup sequence in which two separate read-acquire
1301 * operations read from the same sequence, where the first one
1302 * synchronizes and the other doesn't. Essentially, we can't
1303 * allow any writes to insert themselves between 'release' and
1307 /* Must synchronize */
1308 if (!acquire->synchronize_with(release)) {
1309 set_bad_synchronization();
1312 /* Re-check all pending release sequences */
1313 work_queue->push_back(CheckRelSeqWorkEntry(NULL));
1314 /* Re-check act for mo_graph edges */
1315 work_queue->push_back(MOEdgeWorkEntry(acquire));
1317 /* propagate synchronization to later actions */
1318 action_list_t::reverse_iterator rit = action_trace->rbegin();
1319 for (; (*rit) != acquire; rit++) {
1320 ModelAction *propagate = *rit;
1321 if (acquire->happens_before(propagate)) {
1322 propagate->synchronize_with(acquire);
1323 /* Re-check 'propagate' for mo_graph edges */
1324 work_queue->push_back(MOEdgeWorkEntry(propagate));
1328 /* Break release sequence with new edges:
1329 * release --mo--> write --mo--> rf */
1330 mo_graph->addEdge(release, write);
1331 mo_graph->addEdge(write, rf);
1334 /* See if we have realized a data race */
1339 * Initialize the current action by performing one or more of the following
1340 * actions, as appropriate: merging RMWR and RMWC/RMW actions, stepping forward
1341 * in the NodeStack, manipulating backtracking sets, allocating and
1342 * initializing clock vectors, and computing the promises to fulfill.
1344 * @param curr The current action, as passed from the user context; may be
1345 * freed/invalidated after the execution of this function, with a different
1346 * action "returned" its place (pass-by-reference)
1347 * @return True if curr is a newly-explored action; false otherwise
1349 bool ModelChecker::initialize_curr_action(ModelAction **curr)
1351 ModelAction *newcurr;
1353 if ((*curr)->is_rmwc() || (*curr)->is_rmw()) {
1354 newcurr = process_rmw(*curr);
1357 if (newcurr->is_rmw())
1358 compute_promises(newcurr);
1364 (*curr)->set_seq_number(get_next_seq_num());
1366 newcurr = node_stack->explore_action(*curr, scheduler->get_enabled_array());
1368 /* First restore type and order in case of RMW operation */
1369 if ((*curr)->is_rmwr())
1370 newcurr->copy_typeandorder(*curr);
1372 ASSERT((*curr)->get_location() == newcurr->get_location());
1373 newcurr->copy_from_new(*curr);
1375 /* Discard duplicate ModelAction; use action from NodeStack */
1378 /* Always compute new clock vector */
1379 newcurr->create_cv(get_parent_action(newcurr->get_tid()));
1382 return false; /* Action was explored previously */
1386 /* Always compute new clock vector */
1387 newcurr->create_cv(get_parent_action(newcurr->get_tid()));
1389 /* Assign most recent release fence */
1390 newcurr->set_last_fence_release(get_last_fence_release(newcurr->get_tid()));
1393 * Perform one-time actions when pushing new ModelAction onto
1396 if (newcurr->is_write())
1397 compute_promises(newcurr);
1398 else if (newcurr->is_relseq_fixup())
1399 compute_relseq_breakwrites(newcurr);
1400 else if (newcurr->is_wait())
1401 newcurr->get_node()->set_misc_max(2);
1402 else if (newcurr->is_notify_one()) {
1403 newcurr->get_node()->set_misc_max(get_safe_ptr_action(condvar_waiters_map, newcurr->get_location())->size());
1405 return true; /* This was a new ModelAction */
1410 * @brief Establish reads-from relation between two actions
1412 * Perform basic operations involved with establishing a concrete rf relation,
1413 * including setting the ModelAction data and checking for release sequences.
1415 * @param act The action that is reading (must be a read)
1416 * @param rf The action from which we are reading (must be a write)
1418 * @return True if this read established synchronization
1420 bool ModelChecker::read_from(ModelAction *act, const ModelAction *rf)
1423 ASSERT(rf->is_write());
1425 act->set_read_from(rf);
1426 if (act->is_acquire()) {
1427 rel_heads_list_t release_heads;
1428 get_release_seq_heads(act, act, &release_heads);
1429 int num_heads = release_heads.size();
1430 for (unsigned int i = 0; i < release_heads.size(); i++)
1431 if (!act->synchronize_with(release_heads[i])) {
1432 set_bad_synchronization();
1435 return num_heads > 0;
1441 * Check promises and eliminate potentially-satisfying threads when a thread is
1442 * blocked (e.g., join, lock). A thread which is waiting on another thread can
1443 * no longer satisfy a promise generated from that thread.
1445 * @param blocker The thread on which a thread is waiting
1446 * @param waiting The waiting thread
1448 void ModelChecker::thread_blocking_check_promises(Thread *blocker, Thread *waiting)
1450 for (unsigned int i = 0; i < promises->size(); i++) {
1451 Promise *promise = (*promises)[i];
1452 if (!promise->thread_is_available(waiting->get_id()))
1454 for (unsigned int j = 0; j < promise->get_num_readers(); j++) {
1455 ModelAction *reader = promise->get_reader(j);
1456 if (reader->get_tid() != blocker->get_id())
1458 if (promise->eliminate_thread(waiting->get_id())) {
1459 /* Promise has failed */
1460 priv->failed_promise = true;
1462 /* Only eliminate the 'waiting' thread once */
1470 * @brief Check whether a model action is enabled.
1472 * Checks whether a lock or join operation would be successful (i.e., is the
1473 * lock already locked, or is the joined thread already complete). If not, put
1474 * the action in a waiter list.
1476 * @param curr is the ModelAction to check whether it is enabled.
1477 * @return a bool that indicates whether the action is enabled.
1479 bool ModelChecker::check_action_enabled(ModelAction *curr) {
1480 if (curr->is_lock()) {
1481 std::mutex *lock = curr->get_mutex();
1482 struct std::mutex_state *state = lock->get_state();
1485 } else if (curr->is_thread_join()) {
1486 Thread *blocking = curr->get_thread_operand();
1487 if (!blocking->is_complete()) {
1488 thread_blocking_check_promises(blocking, get_thread(curr));
1497 * This is the heart of the model checker routine. It performs model-checking
1498 * actions corresponding to a given "current action." Among other processes, it
1499 * calculates reads-from relationships, updates synchronization clock vectors,
1500 * forms a memory_order constraints graph, and handles replay/backtrack
1501 * execution when running permutations of previously-observed executions.
1503 * @param curr The current action to process
1504 * @return The ModelAction that is actually executed; may be different than
1505 * curr; may be NULL, if the current action is not enabled to run
1507 ModelAction * ModelChecker::check_current_action(ModelAction *curr)
1510 bool second_part_of_rmw = curr->is_rmwc() || curr->is_rmw();
1511 bool newly_explored = initialize_curr_action(&curr);
1515 wake_up_sleeping_actions(curr);
1517 /* Compute fairness information for CHESS yield algorithm */
1518 if (model->params.yieldon) {
1519 curr->get_node()->update_yield(scheduler);
1522 /* Add the action to lists before any other model-checking tasks */
1523 if (!second_part_of_rmw)
1524 add_action_to_lists(curr);
1526 /* Build may_read_from set for newly-created actions */
1527 if (newly_explored && curr->is_read())
1528 build_may_read_from(curr);
1530 /* Initialize work_queue with the "current action" work */
1531 work_queue_t work_queue(1, CheckCurrWorkEntry(curr));
1532 while (!work_queue.empty() && !has_asserted()) {
1533 WorkQueueEntry work = work_queue.front();
1534 work_queue.pop_front();
1536 switch (work.type) {
1537 case WORK_CHECK_CURR_ACTION: {
1538 ModelAction *act = work.action;
1539 bool update = false; /* update this location's release seq's */
1540 bool update_all = false; /* update all release seq's */
1542 if (process_thread_action(curr))
1545 if (act->is_read() && !second_part_of_rmw && process_read(act))
1548 if (act->is_write() && process_write(act))
1551 if (act->is_fence() && process_fence(act))
1554 if (act->is_mutex_op() && process_mutex(act))
1557 if (act->is_relseq_fixup())
1558 process_relseq_fixup(curr, &work_queue);
1561 work_queue.push_back(CheckRelSeqWorkEntry(NULL));
1563 work_queue.push_back(CheckRelSeqWorkEntry(act->get_location()));
1566 case WORK_CHECK_RELEASE_SEQ:
1567 resolve_release_sequences(work.location, &work_queue);
1569 case WORK_CHECK_MO_EDGES: {
1570 /** @todo Complete verification of work_queue */
1571 ModelAction *act = work.action;
1572 bool updated = false;
1574 if (act->is_read()) {
1575 const ModelAction *rf = act->get_reads_from();
1576 const Promise *promise = act->get_reads_from_promise();
1578 if (r_modification_order(act, rf))
1580 } else if (promise) {
1581 if (r_modification_order(act, promise))
1585 if (act->is_write()) {
1586 if (w_modification_order(act, NULL))
1589 mo_graph->commitChanges();
1592 work_queue.push_back(CheckRelSeqWorkEntry(act->get_location()));
1601 check_curr_backtracking(curr);
1602 set_backtracking(curr);
1606 void ModelChecker::check_curr_backtracking(ModelAction *curr)
1608 Node *currnode = curr->get_node();
1609 Node *parnode = currnode->get_parent();
1611 if ((parnode && !parnode->backtrack_empty()) ||
1612 !currnode->misc_empty() ||
1613 !currnode->read_from_empty() ||
1614 !currnode->promise_empty() ||
1615 !currnode->relseq_break_empty()) {
1616 set_latest_backtrack(curr);
1620 bool ModelChecker::promises_expired() const
1622 for (unsigned int i = 0; i < promises->size(); i++) {
1623 Promise *promise = (*promises)[i];
1624 if (promise->get_expiration() < priv->used_sequence_numbers)
1631 * This is the strongest feasibility check available.
1632 * @return whether the current trace (partial or complete) must be a prefix of
1635 bool ModelChecker::isfeasibleprefix() const
1637 return pending_rel_seqs->size() == 0 && is_feasible_prefix_ignore_relseq();
1641 * Print disagnostic information about an infeasible execution
1642 * @param prefix A string to prefix the output with; if NULL, then a default
1643 * message prefix will be provided
1645 void ModelChecker::print_infeasibility(const char *prefix) const
1649 if (mo_graph->checkForCycles())
1650 ptr += sprintf(ptr, "[mo cycle]");
1651 if (priv->failed_promise)
1652 ptr += sprintf(ptr, "[failed promise]");
1653 if (priv->too_many_reads)
1654 ptr += sprintf(ptr, "[too many reads]");
1655 if (priv->no_valid_reads)
1656 ptr += sprintf(ptr, "[no valid reads-from]");
1657 if (priv->bad_synchronization)
1658 ptr += sprintf(ptr, "[bad sw ordering]");
1659 if (promises_expired())
1660 ptr += sprintf(ptr, "[promise expired]");
1661 if (promises->size() != 0)
1662 ptr += sprintf(ptr, "[unresolved promise]");
1664 model_print("%s: %s\n", prefix ? prefix : "Infeasible", buf);
1668 * Returns whether the current completed trace is feasible, except for pending
1669 * release sequences.
1671 bool ModelChecker::is_feasible_prefix_ignore_relseq() const
1673 return !is_infeasible() && promises->size() == 0;
1677 * Check if the current partial trace is infeasible. Does not check any
1678 * end-of-execution flags, which might rule out the execution. Thus, this is
1679 * useful only for ruling an execution as infeasible.
1680 * @return whether the current partial trace is infeasible.
1682 bool ModelChecker::is_infeasible() const
1684 return mo_graph->checkForCycles() ||
1685 priv->no_valid_reads ||
1686 priv->failed_promise ||
1687 priv->too_many_reads ||
1688 priv->bad_synchronization ||
1692 /** Close out a RMWR by converting previous RMWR into a RMW or READ. */
1693 ModelAction * ModelChecker::process_rmw(ModelAction *act) {
1694 ModelAction *lastread = get_last_action(act->get_tid());
1695 lastread->process_rmw(act);
1696 if (act->is_rmw()) {
1697 if (lastread->get_reads_from())
1698 mo_graph->addRMWEdge(lastread->get_reads_from(), lastread);
1700 mo_graph->addRMWEdge(lastread->get_reads_from_promise(), lastread);
1701 mo_graph->commitChanges();
1707 * A helper function for ModelChecker::check_recency, to check if the current
1708 * thread is able to read from a different write/promise for 'params.maxreads'
1709 * number of steps and if that write/promise should become visible (i.e., is
1710 * ordered later in the modification order). This helps model memory liveness.
1712 * @param curr The current action. Must be a read.
1713 * @param rf The write/promise from which we plan to read
1714 * @param other_rf The write/promise from which we may read
1715 * @return True if we were able to read from other_rf for params.maxreads steps
1717 template <typename T, typename U>
1718 bool ModelChecker::should_read_instead(const ModelAction *curr, const T *rf, const U *other_rf) const
1720 /* Need a different write/promise */
1721 if (other_rf->equals(rf))
1724 /* Only look for "newer" writes/promises */
1725 if (!mo_graph->checkReachable(rf, other_rf))
1728 SnapVector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1729 action_list_t *list = &(*thrd_lists)[id_to_int(curr->get_tid())];
1730 action_list_t::reverse_iterator rit = list->rbegin();
1731 ASSERT((*rit) == curr);
1732 /* Skip past curr */
1735 /* Does this write/promise work for everyone? */
1736 for (int i = 0; i < params.maxreads; i++, rit++) {
1737 ModelAction *act = *rit;
1738 if (!act->may_read_from(other_rf))
1745 * Checks whether a thread has read from the same write or Promise for too many
1746 * times without seeing the effects of a later write/Promise.
1749 * 1) there must a different write/promise that we could read from,
1750 * 2) we must have read from the same write/promise in excess of maxreads times,
1751 * 3) that other write/promise must have been in the reads_from set for maxreads times, and
1752 * 4) that other write/promise must be mod-ordered after the write/promise we are reading.
1754 * If so, we decide that the execution is no longer feasible.
1756 * @param curr The current action. Must be a read.
1757 * @param rf The ModelAction/Promise from which we might read.
1758 * @return True if the read should succeed; false otherwise
1760 template <typename T>
1761 bool ModelChecker::check_recency(ModelAction *curr, const T *rf) const
1763 if (!params.maxreads)
1766 //NOTE: Next check is just optimization, not really necessary....
1767 if (curr->get_node()->get_read_from_past_size() +
1768 curr->get_node()->get_read_from_promise_size() <= 1)
1771 SnapVector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1772 int tid = id_to_int(curr->get_tid());
1773 ASSERT(tid < (int)thrd_lists->size());
1774 action_list_t *list = &(*thrd_lists)[tid];
1775 action_list_t::reverse_iterator rit = list->rbegin();
1776 ASSERT((*rit) == curr);
1777 /* Skip past curr */
1780 action_list_t::reverse_iterator ritcopy = rit;
1781 /* See if we have enough reads from the same value */
1782 for (int count = 0; count < params.maxreads; ritcopy++, count++) {
1783 if (ritcopy == list->rend())
1785 ModelAction *act = *ritcopy;
1786 if (!act->is_read())
1788 if (act->get_reads_from_promise() && !act->get_reads_from_promise()->equals(rf))
1790 if (act->get_reads_from() && !act->get_reads_from()->equals(rf))
1792 if (act->get_node()->get_read_from_past_size() +
1793 act->get_node()->get_read_from_promise_size() <= 1)
1796 for (int i = 0; i < curr->get_node()->get_read_from_past_size(); i++) {
1797 const ModelAction *write = curr->get_node()->get_read_from_past(i);
1798 if (should_read_instead(curr, rf, write))
1799 return false; /* liveness failure */
1801 for (int i = 0; i < curr->get_node()->get_read_from_promise_size(); i++) {
1802 const Promise *promise = curr->get_node()->get_read_from_promise(i);
1803 if (should_read_instead(curr, rf, promise))
1804 return false; /* liveness failure */
1810 * @brief Updates the mo_graph with the constraints imposed from the current
1813 * Basic idea is the following: Go through each other thread and find
1814 * the last action that happened before our read. Two cases:
1816 * -# The action is a write: that write must either occur before
1817 * the write we read from or be the write we read from.
1818 * -# The action is a read: the write that that action read from
1819 * must occur before the write we read from or be the same write.
1821 * @param curr The current action. Must be a read.
1822 * @param rf The ModelAction or Promise that curr reads from. Must be a write.
1823 * @return True if modification order edges were added; false otherwise
1825 template <typename rf_type>
1826 bool ModelChecker::r_modification_order(ModelAction *curr, const rf_type *rf)
1828 SnapVector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1831 ASSERT(curr->is_read());
1833 /* Last SC fence in the current thread */
1834 ModelAction *last_sc_fence_local = get_last_seq_cst_fence(curr->get_tid(), NULL);
1835 ModelAction *last_sc_write = NULL;
1836 if (curr->is_seqcst())
1837 last_sc_write = get_last_seq_cst_write(curr);
1839 /* Iterate over all threads */
1840 for (i = 0; i < thrd_lists->size(); i++) {
1841 /* Last SC fence in thread i */
1842 ModelAction *last_sc_fence_thread_local = NULL;
1843 if (int_to_id((int)i) != curr->get_tid())
1844 last_sc_fence_thread_local = get_last_seq_cst_fence(int_to_id(i), NULL);
1846 /* Last SC fence in thread i, before last SC fence in current thread */
1847 ModelAction *last_sc_fence_thread_before = NULL;
1848 if (last_sc_fence_local)
1849 last_sc_fence_thread_before = get_last_seq_cst_fence(int_to_id(i), last_sc_fence_local);
1851 /* Iterate over actions in thread, starting from most recent */
1852 action_list_t *list = &(*thrd_lists)[i];
1853 action_list_t::reverse_iterator rit;
1854 for (rit = list->rbegin(); rit != list->rend(); rit++) {
1855 ModelAction *act = *rit;
1860 /* Don't want to add reflexive edges on 'rf' */
1861 if (act->equals(rf)) {
1862 if (act->happens_before(curr))
1868 if (act->is_write()) {
1869 /* C++, Section 29.3 statement 5 */
1870 if (curr->is_seqcst() && last_sc_fence_thread_local &&
1871 *act < *last_sc_fence_thread_local) {
1872 added = mo_graph->addEdge(act, rf) || added;
1875 /* C++, Section 29.3 statement 4 */
1876 else if (act->is_seqcst() && last_sc_fence_local &&
1877 *act < *last_sc_fence_local) {
1878 added = mo_graph->addEdge(act, rf) || added;
1881 /* C++, Section 29.3 statement 6 */
1882 else if (last_sc_fence_thread_before &&
1883 *act < *last_sc_fence_thread_before) {
1884 added = mo_graph->addEdge(act, rf) || added;
1889 /* C++, Section 29.3 statement 3 (second subpoint) */
1890 if (curr->is_seqcst() && last_sc_write && act == last_sc_write) {
1891 added = mo_graph->addEdge(act, rf) || added;
1896 * Include at most one act per-thread that "happens
1899 if (act->happens_before(curr)) {
1900 if (act->is_write()) {
1901 added = mo_graph->addEdge(act, rf) || added;
1903 const ModelAction *prevrf = act->get_reads_from();
1904 const Promise *prevrf_promise = act->get_reads_from_promise();
1906 if (!prevrf->equals(rf))
1907 added = mo_graph->addEdge(prevrf, rf) || added;
1908 } else if (!prevrf_promise->equals(rf)) {
1909 added = mo_graph->addEdge(prevrf_promise, rf) || added;
1918 * All compatible, thread-exclusive promises must be ordered after any
1919 * concrete loads from the same thread
1921 for (unsigned int i = 0; i < promises->size(); i++)
1922 if ((*promises)[i]->is_compatible_exclusive(curr))
1923 added = mo_graph->addEdge(rf, (*promises)[i]) || added;
1929 * Updates the mo_graph with the constraints imposed from the current write.
1931 * Basic idea is the following: Go through each other thread and find
1932 * the lastest action that happened before our write. Two cases:
1934 * (1) The action is a write => that write must occur before
1937 * (2) The action is a read => the write that that action read from
1938 * must occur before the current write.
1940 * This method also handles two other issues:
1942 * (I) Sequential Consistency: Making sure that if the current write is
1943 * seq_cst, that it occurs after the previous seq_cst write.
1945 * (II) Sending the write back to non-synchronizing reads.
1947 * @param curr The current action. Must be a write.
1948 * @param send_fv A vector for stashing reads to which we may pass our future
1949 * value. If NULL, then don't record any future values.
1950 * @return True if modification order edges were added; false otherwise
1952 bool ModelChecker::w_modification_order(ModelAction *curr, ModelVector<ModelAction *> *send_fv)
1954 SnapVector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
1957 ASSERT(curr->is_write());
1959 if (curr->is_seqcst()) {
1960 /* We have to at least see the last sequentially consistent write,
1961 so we are initialized. */
1962 ModelAction *last_seq_cst = get_last_seq_cst_write(curr);
1963 if (last_seq_cst != NULL) {
1964 added = mo_graph->addEdge(last_seq_cst, curr) || added;
1968 /* Last SC fence in the current thread */
1969 ModelAction *last_sc_fence_local = get_last_seq_cst_fence(curr->get_tid(), NULL);
1971 /* Iterate over all threads */
1972 for (i = 0; i < thrd_lists->size(); i++) {
1973 /* Last SC fence in thread i, before last SC fence in current thread */
1974 ModelAction *last_sc_fence_thread_before = NULL;
1975 if (last_sc_fence_local && int_to_id((int)i) != curr->get_tid())
1976 last_sc_fence_thread_before = get_last_seq_cst_fence(int_to_id(i), last_sc_fence_local);
1978 /* Iterate over actions in thread, starting from most recent */
1979 action_list_t *list = &(*thrd_lists)[i];
1980 action_list_t::reverse_iterator rit;
1981 for (rit = list->rbegin(); rit != list->rend(); rit++) {
1982 ModelAction *act = *rit;
1985 * 1) If RMW and it actually read from something, then we
1986 * already have all relevant edges, so just skip to next
1989 * 2) If RMW and it didn't read from anything, we should
1990 * whatever edge we can get to speed up convergence.
1992 * 3) If normal write, we need to look at earlier actions, so
1993 * continue processing list.
1995 if (curr->is_rmw()) {
1996 if (curr->get_reads_from() != NULL)
2004 /* C++, Section 29.3 statement 7 */
2005 if (last_sc_fence_thread_before && act->is_write() &&
2006 *act < *last_sc_fence_thread_before) {
2007 added = mo_graph->addEdge(act, curr) || added;
2012 * Include at most one act per-thread that "happens
2015 if (act->happens_before(curr)) {
2017 * Note: if act is RMW, just add edge:
2019 * The following edge should be handled elsewhere:
2020 * readfrom(act) --mo--> act
2022 if (act->is_write())
2023 added = mo_graph->addEdge(act, curr) || added;
2024 else if (act->is_read()) {
2025 //if previous read accessed a null, just keep going
2026 if (act->get_reads_from() == NULL)
2028 added = mo_graph->addEdge(act->get_reads_from(), curr) || added;
2031 } else if (act->is_read() && !act->could_synchronize_with(curr) &&
2032 !act->same_thread(curr)) {
2033 /* We have an action that:
2034 (1) did not happen before us
2035 (2) is a read and we are a write
2036 (3) cannot synchronize with us
2037 (4) is in a different thread
2039 that read could potentially read from our write. Note that
2040 these checks are overly conservative at this point, we'll
2041 do more checks before actually removing the
2045 if (send_fv && thin_air_constraint_may_allow(curr, act)) {
2046 if (!is_infeasible())
2047 send_fv->push_back(act);
2048 else if (curr->is_rmw() && act->is_rmw() && curr->get_reads_from() && curr->get_reads_from() == act->get_reads_from())
2049 add_future_value(curr, act);
2056 * All compatible, thread-exclusive promises must be ordered after any
2057 * concrete stores to the same thread, or else they can be merged with
2060 for (unsigned int i = 0; i < promises->size(); i++)
2061 if ((*promises)[i]->is_compatible_exclusive(curr))
2062 added = mo_graph->addEdge(curr, (*promises)[i]) || added;
2067 /** Arbitrary reads from the future are not allowed. Section 29.3
2068 * part 9 places some constraints. This method checks one result of constraint
2069 * constraint. Others require compiler support. */
2070 bool ModelChecker::thin_air_constraint_may_allow(const ModelAction *writer, const ModelAction *reader) const
2072 if (!writer->is_rmw())
2075 if (!reader->is_rmw())
2078 for (const ModelAction *search = writer->get_reads_from(); search != NULL; search = search->get_reads_from()) {
2079 if (search == reader)
2081 if (search->get_tid() == reader->get_tid() &&
2082 search->happens_before(reader))
2090 * Arbitrary reads from the future are not allowed. Section 29.3 part 9 places
2091 * some constraints. This method checks one the following constraint (others
2092 * require compiler support):
2094 * If X --hb-> Y --mo-> Z, then X should not read from Z.
2096 bool ModelChecker::mo_may_allow(const ModelAction *writer, const ModelAction *reader)
2098 SnapVector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, reader->get_location());
2100 /* Iterate over all threads */
2101 for (i = 0; i < thrd_lists->size(); i++) {
2102 const ModelAction *write_after_read = NULL;
2104 /* Iterate over actions in thread, starting from most recent */
2105 action_list_t *list = &(*thrd_lists)[i];
2106 action_list_t::reverse_iterator rit;
2107 for (rit = list->rbegin(); rit != list->rend(); rit++) {
2108 ModelAction *act = *rit;
2110 /* Don't disallow due to act == reader */
2111 if (!reader->happens_before(act) || reader == act)
2113 else if (act->is_write())
2114 write_after_read = act;
2115 else if (act->is_read() && act->get_reads_from() != NULL)
2116 write_after_read = act->get_reads_from();
2119 if (write_after_read && write_after_read != writer && mo_graph->checkReachable(write_after_read, writer))
2126 * Finds the head(s) of the release sequence(s) containing a given ModelAction.
2127 * The ModelAction under consideration is expected to be taking part in
2128 * release/acquire synchronization as an object of the "reads from" relation.
2129 * Note that this can only provide release sequence support for RMW chains
2130 * which do not read from the future, as those actions cannot be traced until
2131 * their "promise" is fulfilled. Similarly, we may not even establish the
2132 * presence of a release sequence with certainty, as some modification order
2133 * constraints may be decided further in the future. Thus, this function
2134 * "returns" two pieces of data: a pass-by-reference vector of @a release_heads
2135 * and a boolean representing certainty.
2137 * @param rf The action that might be part of a release sequence. Must be a
2139 * @param release_heads A pass-by-reference style return parameter. After
2140 * execution of this function, release_heads will contain the heads of all the
2141 * relevant release sequences, if any exists with certainty
2142 * @param pending A pass-by-reference style return parameter which is only used
2143 * when returning false (i.e., uncertain). Returns most information regarding
2144 * an uncertain release sequence, including any write operations that might
2145 * break the sequence.
2146 * @return true, if the ModelChecker is certain that release_heads is complete;
2149 bool ModelChecker::release_seq_heads(const ModelAction *rf,
2150 rel_heads_list_t *release_heads,
2151 struct release_seq *pending) const
2153 /* Only check for release sequences if there are no cycles */
2154 if (mo_graph->checkForCycles())
2157 for ( ; rf != NULL; rf = rf->get_reads_from()) {
2158 ASSERT(rf->is_write());
2160 if (rf->is_release())
2161 release_heads->push_back(rf);
2162 else if (rf->get_last_fence_release())
2163 release_heads->push_back(rf->get_last_fence_release());
2165 break; /* End of RMW chain */
2167 /** @todo Need to be smarter here... In the linux lock
2168 * example, this will run to the beginning of the program for
2170 /** @todo The way to be smarter here is to keep going until 1
2171 * thread has a release preceded by an acquire and you've seen
2174 /* acq_rel RMW is a sufficient stopping condition */
2175 if (rf->is_acquire() && rf->is_release())
2176 return true; /* complete */
2179 /* read from future: need to settle this later */
2181 return false; /* incomplete */
2184 if (rf->is_release())
2185 return true; /* complete */
2187 /* else relaxed write
2188 * - check for fence-release in the same thread (29.8, stmt. 3)
2189 * - check modification order for contiguous subsequence
2190 * -> rf must be same thread as release */
2192 const ModelAction *fence_release = rf->get_last_fence_release();
2193 /* Synchronize with a fence-release unconditionally; we don't need to
2194 * find any more "contiguous subsequence..." for it */
2196 release_heads->push_back(fence_release);
2198 int tid = id_to_int(rf->get_tid());
2199 SnapVector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, rf->get_location());
2200 action_list_t *list = &(*thrd_lists)[tid];
2201 action_list_t::const_reverse_iterator rit;
2203 /* Find rf in the thread list */
2204 rit = std::find(list->rbegin(), list->rend(), rf);
2205 ASSERT(rit != list->rend());
2207 /* Find the last {write,fence}-release */
2208 for (; rit != list->rend(); rit++) {
2209 if (fence_release && *(*rit) < *fence_release)
2211 if ((*rit)->is_release())
2214 if (rit == list->rend()) {
2215 /* No write-release in this thread */
2216 return true; /* complete */
2217 } else if (fence_release && *(*rit) < *fence_release) {
2218 /* The fence-release is more recent (and so, "stronger") than
2219 * the most recent write-release */
2220 return true; /* complete */
2221 } /* else, need to establish contiguous release sequence */
2222 ModelAction *release = *rit;
2224 ASSERT(rf->same_thread(release));
2226 pending->writes.clear();
2228 bool certain = true;
2229 for (unsigned int i = 0; i < thrd_lists->size(); i++) {
2230 if (id_to_int(rf->get_tid()) == (int)i)
2232 list = &(*thrd_lists)[i];
2234 /* Can we ensure no future writes from this thread may break
2235 * the release seq? */
2236 bool future_ordered = false;
2238 ModelAction *last = get_last_action(int_to_id(i));
2239 Thread *th = get_thread(int_to_id(i));
2240 if ((last && rf->happens_before(last)) ||
2243 future_ordered = true;
2245 ASSERT(!th->is_model_thread() || future_ordered);
2247 for (rit = list->rbegin(); rit != list->rend(); rit++) {
2248 const ModelAction *act = *rit;
2249 /* Reach synchronization -> this thread is complete */
2250 if (act->happens_before(release))
2252 if (rf->happens_before(act)) {
2253 future_ordered = true;
2257 /* Only non-RMW writes can break release sequences */
2258 if (!act->is_write() || act->is_rmw())
2261 /* Check modification order */
2262 if (mo_graph->checkReachable(rf, act)) {
2263 /* rf --mo--> act */
2264 future_ordered = true;
2267 if (mo_graph->checkReachable(act, release))
2268 /* act --mo--> release */
2270 if (mo_graph->checkReachable(release, act) &&
2271 mo_graph->checkReachable(act, rf)) {
2272 /* release --mo-> act --mo--> rf */
2273 return true; /* complete */
2275 /* act may break release sequence */
2276 pending->writes.push_back(act);
2279 if (!future_ordered)
2280 certain = false; /* This thread is uncertain */
2284 release_heads->push_back(release);
2285 pending->writes.clear();
2287 pending->release = release;
2294 * An interface for getting the release sequence head(s) with which a
2295 * given ModelAction must synchronize. This function only returns a non-empty
2296 * result when it can locate a release sequence head with certainty. Otherwise,
2297 * it may mark the internal state of the ModelChecker so that it will handle
2298 * the release sequence at a later time, causing @a acquire to update its
2299 * synchronization at some later point in execution.
2301 * @param acquire The 'acquire' action that may synchronize with a release
2303 * @param read The read action that may read from a release sequence; this may
2304 * be the same as acquire, or else an earlier action in the same thread (i.e.,
2305 * when 'acquire' is a fence-acquire)
2306 * @param release_heads A pass-by-reference return parameter. Will be filled
2307 * with the head(s) of the release sequence(s), if they exists with certainty.
2308 * @see ModelChecker::release_seq_heads
2310 void ModelChecker::get_release_seq_heads(ModelAction *acquire,
2311 ModelAction *read, rel_heads_list_t *release_heads)
2313 const ModelAction *rf = read->get_reads_from();
2314 struct release_seq *sequence = (struct release_seq *)snapshot_calloc(1, sizeof(struct release_seq));
2315 sequence->acquire = acquire;
2316 sequence->read = read;
2318 if (!release_seq_heads(rf, release_heads, sequence)) {
2319 /* add act to 'lazy checking' list */
2320 pending_rel_seqs->push_back(sequence);
2322 snapshot_free(sequence);
2327 * Attempt to resolve all stashed operations that might synchronize with a
2328 * release sequence for a given location. This implements the "lazy" portion of
2329 * determining whether or not a release sequence was contiguous, since not all
2330 * modification order information is present at the time an action occurs.
2332 * @param location The location/object that should be checked for release
2333 * sequence resolutions. A NULL value means to check all locations.
2334 * @param work_queue The work queue to which to add work items as they are
2336 * @return True if any updates occurred (new synchronization, new mo_graph
2339 bool ModelChecker::resolve_release_sequences(void *location, work_queue_t *work_queue)
2341 bool updated = false;
2342 SnapVector<struct release_seq *>::iterator it = pending_rel_seqs->begin();
2343 while (it != pending_rel_seqs->end()) {
2344 struct release_seq *pending = *it;
2345 ModelAction *acquire = pending->acquire;
2346 const ModelAction *read = pending->read;
2348 /* Only resolve sequences on the given location, if provided */
2349 if (location && read->get_location() != location) {
2354 const ModelAction *rf = read->get_reads_from();
2355 rel_heads_list_t release_heads;
2357 complete = release_seq_heads(rf, &release_heads, pending);
2358 for (unsigned int i = 0; i < release_heads.size(); i++) {
2359 if (!acquire->has_synchronized_with(release_heads[i])) {
2360 if (acquire->synchronize_with(release_heads[i]))
2363 set_bad_synchronization();
2368 /* Re-check all pending release sequences */
2369 work_queue->push_back(CheckRelSeqWorkEntry(NULL));
2370 /* Re-check read-acquire for mo_graph edges */
2371 if (acquire->is_read())
2372 work_queue->push_back(MOEdgeWorkEntry(acquire));
2374 /* propagate synchronization to later actions */
2375 action_list_t::reverse_iterator rit = action_trace->rbegin();
2376 for (; (*rit) != acquire; rit++) {
2377 ModelAction *propagate = *rit;
2378 if (acquire->happens_before(propagate)) {
2379 propagate->synchronize_with(acquire);
2380 /* Re-check 'propagate' for mo_graph edges */
2381 work_queue->push_back(MOEdgeWorkEntry(propagate));
2386 it = pending_rel_seqs->erase(it);
2387 snapshot_free(pending);
2393 // If we resolved promises or data races, see if we have realized a data race.
2400 * Performs various bookkeeping operations for the current ModelAction. For
2401 * instance, adds action to the per-object, per-thread action vector and to the
2402 * action trace list of all thread actions.
2404 * @param act is the ModelAction to add.
2406 void ModelChecker::add_action_to_lists(ModelAction *act)
2408 int tid = id_to_int(act->get_tid());
2409 ModelAction *uninit = NULL;
2411 action_list_t *list = get_safe_ptr_action(obj_map, act->get_location());
2412 if (list->empty() && act->is_atomic_var()) {
2413 uninit = get_uninitialized_action(act);
2414 uninit_id = id_to_int(uninit->get_tid());
2415 list->push_front(uninit);
2417 list->push_back(act);
2419 action_trace->push_back(act);
2421 action_trace->push_front(uninit);
2423 SnapVector<action_list_t> *vec = get_safe_ptr_vect_action(obj_thrd_map, act->get_location());
2424 if (tid >= (int)vec->size())
2425 vec->resize(priv->next_thread_id);
2426 (*vec)[tid].push_back(act);
2428 (*vec)[uninit_id].push_front(uninit);
2430 if ((int)thrd_last_action->size() <= tid)
2431 thrd_last_action->resize(get_num_threads());
2432 (*thrd_last_action)[tid] = act;
2434 (*thrd_last_action)[uninit_id] = uninit;
2436 if (act->is_fence() && act->is_release()) {
2437 if ((int)thrd_last_fence_release->size() <= tid)
2438 thrd_last_fence_release->resize(get_num_threads());
2439 (*thrd_last_fence_release)[tid] = act;
2442 if (act->is_wait()) {
2443 void *mutex_loc = (void *) act->get_value();
2444 get_safe_ptr_action(obj_map, mutex_loc)->push_back(act);
2446 SnapVector<action_list_t> *vec = get_safe_ptr_vect_action(obj_thrd_map, mutex_loc);
2447 if (tid >= (int)vec->size())
2448 vec->resize(priv->next_thread_id);
2449 (*vec)[tid].push_back(act);
2454 * @brief Get the last action performed by a particular Thread
2455 * @param tid The thread ID of the Thread in question
2456 * @return The last action in the thread
2458 ModelAction * ModelChecker::get_last_action(thread_id_t tid) const
2460 int threadid = id_to_int(tid);
2461 if (threadid < (int)thrd_last_action->size())
2462 return (*thrd_last_action)[id_to_int(tid)];
2468 * @brief Get the last fence release performed by a particular Thread
2469 * @param tid The thread ID of the Thread in question
2470 * @return The last fence release in the thread, if one exists; NULL otherwise
2472 ModelAction * ModelChecker::get_last_fence_release(thread_id_t tid) const
2474 int threadid = id_to_int(tid);
2475 if (threadid < (int)thrd_last_fence_release->size())
2476 return (*thrd_last_fence_release)[id_to_int(tid)];
2482 * Gets the last memory_order_seq_cst write (in the total global sequence)
2483 * performed on a particular object (i.e., memory location), not including the
2485 * @param curr The current ModelAction; also denotes the object location to
2487 * @return The last seq_cst write
2489 ModelAction * ModelChecker::get_last_seq_cst_write(ModelAction *curr) const
2491 void *location = curr->get_location();
2492 action_list_t *list = get_safe_ptr_action(obj_map, location);
2493 /* Find: max({i in dom(S) | seq_cst(t_i) && isWrite(t_i) && samevar(t_i, t)}) */
2494 action_list_t::reverse_iterator rit;
2495 for (rit = list->rbegin(); (*rit) != curr; rit++)
2497 rit++; /* Skip past curr */
2498 for ( ; rit != list->rend(); rit++)
2499 if ((*rit)->is_write() && (*rit)->is_seqcst())
2505 * Gets the last memory_order_seq_cst fence (in the total global sequence)
2506 * performed in a particular thread, prior to a particular fence.
2507 * @param tid The ID of the thread to check
2508 * @param before_fence The fence from which to begin the search; if NULL, then
2509 * search for the most recent fence in the thread.
2510 * @return The last prior seq_cst fence in the thread, if exists; otherwise, NULL
2512 ModelAction * ModelChecker::get_last_seq_cst_fence(thread_id_t tid, const ModelAction *before_fence) const
2514 /* All fences should have NULL location */
2515 action_list_t *list = get_safe_ptr_action(obj_map, NULL);
2516 action_list_t::reverse_iterator rit = list->rbegin();
2519 for (; rit != list->rend(); rit++)
2520 if (*rit == before_fence)
2523 ASSERT(*rit == before_fence);
2527 for (; rit != list->rend(); rit++)
2528 if ((*rit)->is_fence() && (tid == (*rit)->get_tid()) && (*rit)->is_seqcst())
2534 * Gets the last unlock operation performed on a particular mutex (i.e., memory
2535 * location). This function identifies the mutex according to the current
2536 * action, which is presumed to perform on the same mutex.
2537 * @param curr The current ModelAction; also denotes the object location to
2539 * @return The last unlock operation
2541 ModelAction * ModelChecker::get_last_unlock(ModelAction *curr) const
2543 void *location = curr->get_location();
2544 action_list_t *list = get_safe_ptr_action(obj_map, location);
2545 /* Find: max({i in dom(S) | isUnlock(t_i) && samevar(t_i, t)}) */
2546 action_list_t::reverse_iterator rit;
2547 for (rit = list->rbegin(); rit != list->rend(); rit++)
2548 if ((*rit)->is_unlock() || (*rit)->is_wait())
2553 ModelAction * ModelChecker::get_parent_action(thread_id_t tid) const
2555 ModelAction *parent = get_last_action(tid);
2557 parent = get_thread(tid)->get_creation();
2562 * Returns the clock vector for a given thread.
2563 * @param tid The thread whose clock vector we want
2564 * @return Desired clock vector
2566 ClockVector * ModelChecker::get_cv(thread_id_t tid) const
2568 return get_parent_action(tid)->get_cv();
2572 * @brief Find the promise (if any) to resolve for the current action and
2573 * remove it from the pending promise vector
2574 * @param curr The current ModelAction. Should be a write.
2575 * @return The Promise to resolve, if any; otherwise NULL
2577 Promise * ModelChecker::pop_promise_to_resolve(const ModelAction *curr)
2579 for (unsigned int i = 0; i < promises->size(); i++)
2580 if (curr->get_node()->get_promise(i)) {
2581 Promise *ret = (*promises)[i];
2582 promises->erase(promises->begin() + i);
2589 * Resolve a Promise with a current write.
2590 * @param write The ModelAction that is fulfilling Promises
2591 * @param promise The Promise to resolve
2592 * @return True if the Promise was successfully resolved; false otherwise
2594 bool ModelChecker::resolve_promise(ModelAction *write, Promise *promise)
2596 ModelVector<ModelAction *> actions_to_check;
2598 for (unsigned int i = 0; i < promise->get_num_readers(); i++) {
2599 ModelAction *read = promise->get_reader(i);
2600 read_from(read, write);
2601 actions_to_check.push_back(read);
2603 /* Make sure the promise's value matches the write's value */
2604 ASSERT(promise->is_compatible(write) && promise->same_value(write));
2605 if (!mo_graph->resolvePromise(promise, write))
2606 priv->failed_promise = true;
2609 * @todo It is possible to end up in an inconsistent state, where a
2610 * "resolved" promise may still be referenced if
2611 * CycleGraph::resolvePromise() failed, so don't delete 'promise'.
2613 * Note that the inconsistency only matters when dumping mo_graph to
2619 //Check whether reading these writes has made threads unable to
2621 for (unsigned int i = 0; i < actions_to_check.size(); i++) {
2622 ModelAction *read = actions_to_check[i];
2623 mo_check_promises(read, true);
2630 * Compute the set of promises that could potentially be satisfied by this
2631 * action. Note that the set computation actually appears in the Node, not in
2633 * @param curr The ModelAction that may satisfy promises
2635 void ModelChecker::compute_promises(ModelAction *curr)
2637 for (unsigned int i = 0; i < promises->size(); i++) {
2638 Promise *promise = (*promises)[i];
2639 if (!promise->is_compatible(curr) || !promise->same_value(curr))
2642 bool satisfy = true;
2643 for (unsigned int j = 0; j < promise->get_num_readers(); j++) {
2644 const ModelAction *act = promise->get_reader(j);
2645 if (act->happens_before(curr) ||
2646 act->could_synchronize_with(curr)) {
2652 curr->get_node()->set_promise(i);
2656 /** Checks promises in response to change in ClockVector Threads. */
2657 void ModelChecker::check_promises(thread_id_t tid, ClockVector *old_cv, ClockVector *merge_cv)
2659 for (unsigned int i = 0; i < promises->size(); i++) {
2660 Promise *promise = (*promises)[i];
2661 if (!promise->thread_is_available(tid))
2663 for (unsigned int j = 0; j < promise->get_num_readers(); j++) {
2664 const ModelAction *act = promise->get_reader(j);
2665 if ((!old_cv || !old_cv->synchronized_since(act)) &&
2666 merge_cv->synchronized_since(act)) {
2667 if (promise->eliminate_thread(tid)) {
2668 /* Promise has failed */
2669 priv->failed_promise = true;
2677 void ModelChecker::check_promises_thread_disabled()
2679 for (unsigned int i = 0; i < promises->size(); i++) {
2680 Promise *promise = (*promises)[i];
2681 if (promise->has_failed()) {
2682 priv->failed_promise = true;
2689 * @brief Checks promises in response to addition to modification order for
2692 * We test whether threads are still available for satisfying promises after an
2693 * addition to our modification order constraints. Those that are unavailable
2694 * are "eliminated". Once all threads are eliminated from satisfying a promise,
2695 * that promise has failed.
2697 * @param act The ModelAction which updated the modification order
2698 * @param is_read_check Should be true if act is a read and we must check for
2699 * updates to the store from which it read (there is a distinction here for
2700 * RMW's, which are both a load and a store)
2702 void ModelChecker::mo_check_promises(const ModelAction *act, bool is_read_check)
2704 const ModelAction *write = is_read_check ? act->get_reads_from() : act;
2706 for (unsigned int i = 0; i < promises->size(); i++) {
2707 Promise *promise = (*promises)[i];
2709 // Is this promise on the same location?
2710 if (!promise->same_location(write))
2713 for (unsigned int j = 0; j < promise->get_num_readers(); j++) {
2714 const ModelAction *pread = promise->get_reader(j);
2715 if (!pread->happens_before(act))
2717 if (mo_graph->checkPromise(write, promise)) {
2718 priv->failed_promise = true;
2724 // Don't do any lookups twice for the same thread
2725 if (!promise->thread_is_available(act->get_tid()))
2728 if (mo_graph->checkReachable(promise, write)) {
2729 if (mo_graph->checkPromise(write, promise)) {
2730 priv->failed_promise = true;
2738 * Compute the set of writes that may break the current pending release
2739 * sequence. This information is extracted from previou release sequence
2742 * @param curr The current ModelAction. Must be a release sequence fixup
2745 void ModelChecker::compute_relseq_breakwrites(ModelAction *curr)
2747 if (pending_rel_seqs->empty())
2750 struct release_seq *pending = pending_rel_seqs->back();
2751 for (unsigned int i = 0; i < pending->writes.size(); i++) {
2752 const ModelAction *write = pending->writes[i];
2753 curr->get_node()->add_relseq_break(write);
2756 /* NULL means don't break the sequence; just synchronize */
2757 curr->get_node()->add_relseq_break(NULL);
2761 * Build up an initial set of all past writes that this 'read' action may read
2762 * from, as well as any previously-observed future values that must still be valid.
2764 * @param curr is the current ModelAction that we are exploring; it must be a
2767 void ModelChecker::build_may_read_from(ModelAction *curr)
2769 SnapVector<action_list_t> *thrd_lists = get_safe_ptr_vect_action(obj_thrd_map, curr->get_location());
2771 ASSERT(curr->is_read());
2773 ModelAction *last_sc_write = NULL;
2775 if (curr->is_seqcst())
2776 last_sc_write = get_last_seq_cst_write(curr);
2778 /* Iterate over all threads */
2779 for (i = 0; i < thrd_lists->size(); i++) {
2780 /* Iterate over actions in thread, starting from most recent */
2781 action_list_t *list = &(*thrd_lists)[i];
2782 action_list_t::reverse_iterator rit;
2783 for (rit = list->rbegin(); rit != list->rend(); rit++) {
2784 ModelAction *act = *rit;
2786 /* Only consider 'write' actions */
2787 if (!act->is_write() || act == curr)
2790 /* Don't consider more than one seq_cst write if we are a seq_cst read. */
2791 bool allow_read = true;
2793 if (curr->is_seqcst() && (act->is_seqcst() || (last_sc_write != NULL && act->happens_before(last_sc_write))) && act != last_sc_write)
2795 else if (curr->get_sleep_flag() && !curr->is_seqcst() && !sleep_can_read_from(curr, act))
2799 /* Only add feasible reads */
2800 mo_graph->startChanges();
2801 r_modification_order(curr, act);
2802 if (!is_infeasible())
2803 curr->get_node()->add_read_from_past(act);
2804 mo_graph->rollbackChanges();
2807 /* Include at most one act per-thread that "happens before" curr */
2808 if (act->happens_before(curr))
2813 /* Inherit existing, promised future values */
2814 for (i = 0; i < promises->size(); i++) {
2815 const Promise *promise = (*promises)[i];
2816 const ModelAction *promise_read = promise->get_reader(0);
2817 if (promise_read->same_var(curr)) {
2818 /* Only add feasible future-values */
2819 mo_graph->startChanges();
2820 r_modification_order(curr, promise);
2821 if (!is_infeasible())
2822 curr->get_node()->add_read_from_promise(promise_read);
2823 mo_graph->rollbackChanges();
2827 /* We may find no valid may-read-from only if the execution is doomed */
2828 if (!curr->get_node()->read_from_size()) {
2829 priv->no_valid_reads = true;
2833 if (DBG_ENABLED()) {
2834 model_print("Reached read action:\n");
2836 model_print("Printing read_from_past\n");
2837 curr->get_node()->print_read_from_past();
2838 model_print("End printing read_from_past\n");
2842 bool ModelChecker::sleep_can_read_from(ModelAction *curr, const ModelAction *write)
2844 for ( ; write != NULL; write = write->get_reads_from()) {
2845 /* UNINIT actions don't have a Node, and they never sleep */
2846 if (write->is_uninitialized())
2848 Node *prevnode = write->get_node()->get_parent();
2850 bool thread_sleep = prevnode->enabled_status(curr->get_tid()) == THREAD_SLEEP_SET;
2851 if (write->is_release() && thread_sleep)
2853 if (!write->is_rmw())
2860 * @brief Get an action representing an uninitialized atomic
2862 * This function may create a new one or try to retrieve one from the NodeStack
2864 * @param curr The current action, which prompts the creation of an UNINIT action
2865 * @return A pointer to the UNINIT ModelAction
2867 ModelAction * ModelChecker::get_uninitialized_action(const ModelAction *curr) const
2869 Node *node = curr->get_node();
2870 ModelAction *act = node->get_uninit_action();
2872 act = new ModelAction(ATOMIC_UNINIT, std::memory_order_relaxed, curr->get_location(), model->params.uninitvalue, model_thread);
2873 node->set_uninit_action(act);
2875 act->create_cv(NULL);
2879 static void print_list(action_list_t *list)
2881 action_list_t::iterator it;
2883 model_print("---------------------------------------------------------------------\n");
2885 unsigned int hash = 0;
2887 for (it = list->begin(); it != list->end(); it++) {
2888 const ModelAction *act = *it;
2889 if (act->get_seq_number() > 0)
2891 hash = hash^(hash<<3)^((*it)->hash());
2893 model_print("HASH %u\n", hash);
2894 model_print("---------------------------------------------------------------------\n");
2897 #if SUPPORT_MOD_ORDER_DUMP
2898 void ModelChecker::dumpGraph(char *filename) const
2901 sprintf(buffer, "%s.dot", filename);
2902 FILE *file = fopen(buffer, "w");
2903 fprintf(file, "digraph %s {\n", filename);
2904 mo_graph->dumpNodes(file);
2905 ModelAction **thread_array = (ModelAction **)model_calloc(1, sizeof(ModelAction *) * get_num_threads());
2907 for (action_list_t::iterator it = action_trace->begin(); it != action_trace->end(); it++) {
2908 ModelAction *act = *it;
2909 if (act->is_read()) {
2910 mo_graph->dot_print_node(file, act);
2911 if (act->get_reads_from())
2912 mo_graph->dot_print_edge(file,
2913 act->get_reads_from(),
2915 "label=\"rf\", color=red, weight=2");
2917 mo_graph->dot_print_edge(file,
2918 act->get_reads_from_promise(),
2920 "label=\"rf\", color=red");
2922 if (thread_array[act->get_tid()]) {
2923 mo_graph->dot_print_edge(file,
2924 thread_array[id_to_int(act->get_tid())],
2926 "label=\"sb\", color=blue, weight=400");
2929 thread_array[act->get_tid()] = act;
2931 fprintf(file, "}\n");
2932 model_free(thread_array);
2937 /** @brief Prints an execution trace summary. */
2938 void ModelChecker::print_summary() const
2940 #if SUPPORT_MOD_ORDER_DUMP
2941 char buffername[100];
2942 sprintf(buffername, "exec%04u", stats.num_total);
2943 mo_graph->dumpGraphToFile(buffername);
2944 sprintf(buffername, "graph%04u", stats.num_total);
2945 dumpGraph(buffername);
2948 model_print("Execution %d:", stats.num_total);
2949 if (isfeasibleprefix()) {
2950 if (scheduler->all_threads_sleeping())
2951 model_print(" SLEEP-SET REDUNDANT");
2954 print_infeasibility(" INFEASIBLE");
2955 print_list(action_trace);
2957 if (!promises->empty()) {
2958 model_print("Pending promises:\n");
2959 for (unsigned int i = 0; i < promises->size(); i++) {
2960 model_print(" [P%u] ", i);
2961 (*promises)[i]->print();
2968 * Add a Thread to the system for the first time. Should only be called once
2970 * @param t The Thread to add
2972 void ModelChecker::add_thread(Thread *t)
2974 thread_map->put(id_to_int(t->get_id()), t);
2975 scheduler->add_thread(t);
2979 * @brief Get a Thread reference by its ID
2980 * @param tid The Thread's ID
2981 * @return A Thread reference
2983 Thread * ModelChecker::get_thread(thread_id_t tid) const
2985 return thread_map->get(id_to_int(tid));
2989 * @brief Get a reference to the Thread in which a ModelAction was executed
2990 * @param act The ModelAction
2991 * @return A Thread reference
2993 Thread * ModelChecker::get_thread(const ModelAction *act) const
2995 return get_thread(act->get_tid());
2999 * @brief Get a Promise's "promise number"
3001 * A "promise number" is an index number that is unique to a promise, valid
3002 * only for a specific snapshot of an execution trace. Promises may come and go
3003 * as they are generated an resolved, so an index only retains meaning for the
3006 * @param promise The Promise to check
3007 * @return The promise index, if the promise still is valid; otherwise -1
3009 int ModelChecker::get_promise_number(const Promise *promise) const
3011 for (unsigned int i = 0; i < promises->size(); i++)
3012 if ((*promises)[i] == promise)
3019 * @brief Check if a Thread is currently enabled
3020 * @param t The Thread to check
3021 * @return True if the Thread is currently enabled
3023 bool ModelChecker::is_enabled(Thread *t) const
3025 return scheduler->is_enabled(t);
3029 * @brief Check if a Thread is currently enabled
3030 * @param tid The ID of the Thread to check
3031 * @return True if the Thread is currently enabled
3033 bool ModelChecker::is_enabled(thread_id_t tid) const
3035 return scheduler->is_enabled(tid);
3039 * Switch from a model-checker context to a user-thread context. This is the
3040 * complement of ModelChecker::switch_to_master and must be called from the
3041 * model-checker context
3043 * @param thread The user-thread to switch to
3045 void ModelChecker::switch_from_master(Thread *thread)
3047 scheduler->set_current_thread(thread);
3048 Thread::swap(&system_context, thread);
3052 * Switch from a user-context to the "master thread" context (a.k.a. system
3053 * context). This switch is made with the intention of exploring a particular
3054 * model-checking action (described by a ModelAction object). Must be called
3055 * from a user-thread context.
3057 * @param act The current action that will be explored. May be NULL only if
3058 * trace is exiting via an assertion (see ModelChecker::set_assert and
3059 * ModelChecker::has_asserted).
3060 * @return Return the value returned by the current action
3062 uint64_t ModelChecker::switch_to_master(ModelAction *act)
3065 Thread *old = thread_current();
3066 scheduler->set_current_thread(NULL);
3067 ASSERT(!old->get_pending());
3068 old->set_pending(act);
3069 if (Thread::swap(old, &system_context) < 0) {
3070 perror("swap threads");
3073 return old->get_return_value();
3077 * Takes the next step in the execution, if possible.
3078 * @param curr The current step to take
3079 * @return Returns the next Thread to run, if any; NULL if this execution
3082 Thread * ModelChecker::take_step(ModelAction *curr)
3084 Thread *curr_thrd = get_thread(curr);
3085 ASSERT(curr_thrd->get_state() == THREAD_READY);
3087 ASSERT(check_action_enabled(curr)); /* May have side effects? */
3088 curr = check_current_action(curr);
3091 if (curr_thrd->is_blocked() || curr_thrd->is_complete())
3092 scheduler->remove_thread(curr_thrd);
3094 return action_select_next_thread(curr);
3097 /** Wrapper to run the user's main function, with appropriate arguments */
3098 void user_main_wrapper(void *)
3100 user_main(model->params.argc, model->params.argv);
3103 /** @return True if the execution has taken too many steps */
3104 bool ModelChecker::too_many_steps() const
3106 return params.bound != 0 && priv->used_sequence_numbers > params.bound;
3109 bool ModelChecker::should_terminate_execution()
3111 /* Infeasible -> don't take any more steps */
3112 if (is_infeasible())
3114 else if (isfeasibleprefix() && have_bug_reports()) {
3119 if (too_many_steps())
3124 /** @brief Run ModelChecker for the user program */
3125 void ModelChecker::run()
3129 Thread *t = new Thread(&user_thread, &user_main_wrapper, NULL, NULL);
3134 * Stash next pending action(s) for thread(s). There
3135 * should only need to stash one thread's action--the
3136 * thread which just took a step--plus the first step
3137 * for any newly-created thread
3139 for (unsigned int i = 0; i < get_num_threads(); i++) {
3140 thread_id_t tid = int_to_id(i);
3141 Thread *thr = get_thread(tid);
3142 if (!thr->is_model_thread() && !thr->is_complete() && !thr->get_pending()) {
3143 switch_from_master(thr);
3144 if (thr->is_waiting_on(thr))
3145 assert_bug("Deadlock detected (thread %u)", i);
3149 /* Don't schedule threads which should be disabled */
3150 for (unsigned int i = 0; i < get_num_threads(); i++) {
3151 Thread *th = get_thread(int_to_id(i));
3152 ModelAction *act = th->get_pending();
3153 if (act && is_enabled(th) && !check_action_enabled(act)) {
3154 scheduler->sleep(th);
3158 /* Catch assertions from prior take_step or from
3159 * between-ModelAction bugs (e.g., data races) */
3164 t = get_next_thread();
3165 if (!t || t->is_model_thread())
3168 /* Consume the next action for a Thread */
3169 ModelAction *curr = t->get_pending();
3170 t->set_pending(NULL);
3171 t = take_step(curr);
3172 } while (!should_terminate_execution());
3175 * Launch end-of-execution release sequence fixups only when
3176 * the execution is otherwise feasible AND there are:
3178 * (1) pending release sequences
3179 * (2) pending assertions that could be invalidated by a change
3180 * in clock vectors (i.e., data races)
3181 * (3) no pending promises
3183 while (!pending_rel_seqs->empty() &&
3184 is_feasible_prefix_ignore_relseq() &&
3185 !unrealizedraces.empty()) {
3186 model_print("*** WARNING: release sequence fixup action "
3187 "(%zu pending release seuqence(s)) ***\n",
3188 pending_rel_seqs->size());
3189 ModelAction *fixup = new ModelAction(MODEL_FIXUP_RELSEQ,
3190 std::memory_order_seq_cst, NULL, VALUE_NONE,
3194 } while (next_execution());
3196 model_print("******* Model-checking complete: *******\n");