[OpenMP] libomp: add UNLIKELY hints to rarely executed branches

[AndreyChurbanov]

Summary

Added UNLIKELY hint to one-time or rarely executed branches.
This improves performance of the library on some tasking benchmarks.

[AndreyChurbanov]

Rebased

[tianshilei1992]

LGTM.

openmp/runtime/src/kmp_itt.inl

 
 /* Taskwait reporting.
    ITT need an address (void *) to be specified as a sync object. OpenMP RTL
    does not have taskwait structure, so we need to construct something. */
 
 void *__kmp_itt_taskwait_object(int gtid) {
   void *object = NULL;
 #if USE_ITT_NOTIFY
-  if (__itt_sync_create_ptr) {
+  if (UNLIKELY(__itt_sync_create_ptr)) {
     kmp_info_t *thread = __kmp_thread_from_gtid(gtid);
     kmp_taskdata_t *taskdata = thread->th.th_current_task;
     object = reinterpret_cast<void *>(kmp_uintptr_t(taskdata) +
                                       taskdata->td_taskwait_counter %
                                           sizeof(kmp_taskdata_t));
   }
 #endif
   return object;
/* Task reporting.
    Only those tasks are reported which are executed by a thread spinning at
    barrier (or taskwait). Synch object passed to the function must be barrier of
    taskwait the threads waiting at. */
 
 void __kmp_itt_task_starting(
     void *object // ITT sync object: barrier or taskwait.
     ) {
 #if USE_ITT_NOTIFY
-  if (object != NULL) {
+  if (UNLIKELY(object != NULL)) {
     KMP_ITT_DEBUG_LOCK();
     __itt_sync_cancel(object);
     KMP_ITT_DEBUG_PRINT("[tsk sta] scan( %p )\n", object);
   }
 #endif
 } // __kmp_itt_task_starting
 
 // -----------------------------------------------------------------------------

openmp/runtime/src/kmp_tasking.cpp

if (current->td_flags.tasktype == TASK_EXPLICIT ||
         KMP_DEBUG_ASSERT(parent != NULL);
       }
       if (parent != current)
         return false;
     }
   }
   // Check mutexinoutset dependencies, acquire locks
   kmp_depnode_t *node = tasknew->td_depnode;
-  if (node && (node->dn.mtx_num_locks > 0)) {
+  if (UNLIKELY(node && (node->dn.mtx_num_locks > 0))) {
     for (int i = 0; i < node->dn.mtx_num_locks; ++i) {
       KMP_DEBUG_ASSERT(node->dn.mtx_locks[i] != NULL);
       if (__kmp_test_lock(node->dn.mtx_locks[i], gtid))
         continue;
       // could not get the lock, release previous locks
       for (int j = i - 1; j >= 0; --j)
         __kmp_release_lock(node->dn.mtx_locks[j], gtid);
       return false;
static kmp_int32 __kmp_push_task(kmp_int32 gtid, kmp_task_t *task) {
   kmp_taskdata_t *taskdata = KMP_TASK_TO_TASKDATA(task);
   kmp_task_team_t *task_team = thread->th.th_task_team;
   kmp_int32 tid = __kmp_tid_from_gtid(gtid);
   kmp_thread_data_t *thread_data;
 
   KA_TRACE(20,
            ("__kmp_push_task: T#%d trying to push task %p.\n", gtid, taskdata));
 
-  if (taskdata->td_flags.tiedness == TASK_UNTIED) {
+  if (UNLIKELY(taskdata->td_flags.tiedness == TASK_UNTIED)) {
     // untied task needs to increment counter so that the task structure is not
     // freed prematurely
     kmp_int32 counter = 1 + KMP_ATOMIC_INC(&taskdata->td_untied_count);
     KMP_DEBUG_USE_VAR(counter);
     KA_TRACE(
         20,
         ("__kmp_push_task: T#%d untied_count (%d) incremented for task %p\n",
          gtid, counter, taskdata));
   }
 
   // The first check avoids building task_team thread data if serialized
-  if (taskdata->td_flags.task_serial) {
+  if (UNLIKELY(taskdata->td_flags.task_serial)) {
     KA_TRACE(20, ("__kmp_push_task: T#%d team serialized; returning "
                   "TASK_NOT_PUSHED for task %p\n",
                   gtid, taskdata));
     return TASK_NOT_PUSHED;
   }
 
   // Now that serialized tasks have returned, we can assume that we are not in
   // immediate exec mode
   KMP_DEBUG_ASSERT(__kmp_tasking_mode != tskm_immediate_exec);
-  if (!KMP_TASKING_ENABLED(task_team)) {
+  if (UNLIKELY(!KMP_TASKING_ENABLED(task_team))) {
     __kmp_enable_tasking(task_team, thread);
   }
   KMP_DEBUG_ASSERT(TCR_4(task_team->tt.tt_found_tasks) == TRUE);
   KMP_DEBUG_ASSERT(TCR_PTR(task_team->tt.tt_threads_data) != NULL);
 
   // Find tasking deque specific to encountering thread
   thread_data = &task_team->tt.tt_threads_data[tid];
 
   // No lock needed since only owner can allocate
-  if (thread_data->td.td_deque == NULL) {
+  if (UNLIKELY(thread_data->td.td_deque == NULL)) {
     __kmp_alloc_task_deque(thread, thread_data);
   }
 
   int locked = 0;
   // Check if deque is full
   if (TCR_4(thread_data->td.td_deque_ntasks) >=
       TASK_DEQUE_SIZE(thread_data->td)) {
     if (__kmp_enable_task_throttling &&

 
 // Pop task from stack if tied
 #ifdef BUILD_TIED_TASK_STACK
   if (taskdata->td_flags.tiedness == TASK_TIED) {
     __kmp_pop_task_stack(gtid, thread, taskdata);
   }
 #endif /* BUILD_TIED_TASK_STACK */
 
-  if (taskdata->td_flags.tiedness == TASK_UNTIED) {
+  if (UNLIKELY(taskdata->td_flags.tiedness == TASK_UNTIED)) {
     // untied task needs to check the counter so that the task structure is not
     // freed prematurely
     kmp_int32 counter = KMP_ATOMIC_DEC(&taskdata->td_untied_count) - 1;
     KA_TRACE(
         20,
         ("__kmp_task_finish: T#%d untied_count (%d) decremented for task %p\n",
          gtid, counter, taskdata));
     if (counter > 0) {
kmp_task_t *__kmp_task_alloc(ident_t *loc_ref, kmp_int32 gtid,
                              kmp_routine_entry_t task_entry) {
   kmp_task_t *task;
   kmp_taskdata_t *taskdata;
   kmp_info_t *thread = __kmp_threads[gtid];
   kmp_team_t *team = thread->th.th_team;
   kmp_taskdata_t *parent_task = thread->th.th_current_task;
   size_t shareds_offset;
 
-  if (!TCR_4(__kmp_init_middle))
+  if (UNLIKELY(!TCR_4(__kmp_init_middle)))
     __kmp_middle_initialize();
 
   KA_TRACE(10, ("__kmp_task_alloc(enter): T#%d loc=%p, flags=(0x%x) "
                 "sizeof_task=%ld sizeof_shared=%ld entry=%p\n",
                 gtid, loc_ref, *((kmp_int32 *)flags), sizeof_kmp_task_t,
                 sizeof_shareds, task_entry));
 
   if (parent_task->td_flags.final) {
static void __kmp_invoke_task(kmp_int32 gtid, kmp_task_t *task,
                               kmp_taskdata_t *current_task) {
   kmp_taskdata_t *taskdata = KMP_TASK_TO_TASKDATA(task);
   kmp_info_t *thread;
   int discard = 0 /* false */;
   KA_TRACE(
       30, ("__kmp_invoke_task(enter): T#%d invoking task %p, current_task=%p\n",
            gtid, taskdata, current_task));
   KMP_DEBUG_ASSERT(task);
-  if (taskdata->td_flags.proxy == TASK_PROXY &&
-      taskdata->td_flags.complete == 1) {
+  if (UNLIKELY(taskdata->td_flags.proxy == TASK_PROXY &&
+               taskdata->td_flags.complete == 1)) {
     // This is a proxy task that was already completed but it needs to run
     // its bottom-half finish
     KA_TRACE(
         30,
         ("__kmp_invoke_task: T#%d running bottom finish for proxy task %p\n",
          gtid, taskdata));
 
     __kmp_bottom_half_finish_proxy(gtid, task);
#endif
   if (taskdata->td_flags.proxy != TASK_PROXY) {
     ANNOTATE_HAPPENS_AFTER(task);
     __kmp_task_start(gtid, task, current_task); // OMPT only if not discarded
   }
 
   // TODO: cancel tasks if the parallel region has also been cancelled
   // TODO: check if this sequence can be hoisted above __kmp_task_start
   // if cancellation has been enabled for this run ...
-  if (__kmp_omp_cancellation) {
+  if (UNLIKELY(__kmp_omp_cancellation)) {
     thread = __kmp_threads[gtid];
     kmp_team_t *this_team = thread->th.th_team;
     kmp_taskgroup_t *taskgroup = taskdata->td_taskgroup;
     if ((taskgroup && taskgroup->cancel_request) ||
         (this_team->t.t_cancel_request == cancel_parallel)) {
 #if OMPT_SUPPORT && OMPT_OPTIONAL
       ompt_data_t *task_data;
       if (UNLIKELY(ompt_enabled.ompt_callback_cancel)) {

 // Note: These values are used by ITT events as well.
 #endif /* USE_ITT_BUILD */
     taskdata->td_taskwait_counter += 1;
     taskdata->td_taskwait_ident = loc_ref;
     taskdata->td_taskwait_thread = gtid + 1;
 
 #if USE_ITT_BUILD
     void *itt_sync_obj = __kmp_itt_taskwait_object(gtid);
-    if (itt_sync_obj != NULL)
+    if (UNLIKELY(itt_sync_obj != NULL))
       __kmp_itt_taskwait_starting(gtid, itt_sync_obj);
 #endif /* USE_ITT_BUILD */
 
     bool must_wait =
         !taskdata->td_flags.team_serial && !taskdata->td_flags.final;
 
     must_wait = must_wait || (thread->th.th_task_team != NULL &&
                               thread->th.th_task_team->tt.tt_found_proxy_tasks);
     if (must_wait) {
       kmp_flag_32 flag(RCAST(std::atomic<kmp_uint32> *,
                              &(taskdata->td_incomplete_child_tasks)),
                        0U);
       while (KMP_ATOMIC_LD_ACQ(&taskdata->td_incomplete_child_tasks) != 0) {
         flag.execute_tasks(thread, gtid, FALSE,
                            &thread_finished USE_ITT_BUILD_ARG(itt_sync_obj),
                            __kmp_task_stealing_constraint);
       }
     }
 #if USE_ITT_BUILD
-    if (itt_sync_obj != NULL)
+    if (UNLIKELY(itt_sync_obj != NULL))
       __kmp_itt_taskwait_finished(gtid, itt_sync_obj);
     KMP_FSYNC_ACQUIRED(taskdata); // acquire self - sync with children
 #endif /* USE_ITT_BUILD */
 
     // Debugger:  The taskwait is completed. Location remains, but thread is
     // negated.
     taskdata->td_taskwait_thread = -taskdata->td_taskwait_thread;
 

 // Note: These values are used by ITT events as well.
 #endif /* USE_ITT_BUILD */
     taskdata->td_taskwait_counter += 1;
     taskdata->td_taskwait_ident = loc_ref;
     taskdata->td_taskwait_thread = gtid + 1;
 
 #if USE_ITT_BUILD
     void *itt_sync_obj = __kmp_itt_taskwait_object(gtid);
-    if (itt_sync_obj != NULL)
+    if (UNLIKELY(itt_sync_obj != NULL))
       __kmp_itt_taskwait_starting(gtid, itt_sync_obj);
 #endif /* USE_ITT_BUILD */
     if (!taskdata->td_flags.team_serial) {
       kmp_task_team_t *task_team = thread->th.th_task_team;
       if (task_team != NULL) {
         if (KMP_TASKING_ENABLED(task_team)) {
 #if OMPT_SUPPORT
           if (UNLIKELY(ompt_enabled.enabled))
             thread->th.ompt_thread_info.ompt_task_yielded = 1;
 #endif
           __kmp_execute_tasks_32(
               thread, gtid, NULL, FALSE,
               &thread_finished USE_ITT_BUILD_ARG(itt_sync_obj),
               __kmp_task_stealing_constraint);
 #if OMPT_SUPPORT
           if (UNLIKELY(ompt_enabled.enabled))
             thread->th.ompt_thread_info.ompt_task_yielded = 0;
 #endif
         }
       }
     }
 #if USE_ITT_BUILD
-    if (itt_sync_obj != NULL)
+    if (UNLIKELY(itt_sync_obj != NULL))
       __kmp_itt_taskwait_finished(gtid, itt_sync_obj);
 #endif /* USE_ITT_BUILD */
 
     // Debugger:  The taskwait is completed. Location remains, but thread is
     // negated.
     taskdata->td_taskwait_thread = -taskdata->td_taskwait_thread;
   }
 
if (__kmp_tasking_mode != tskm_immediate_exec) {
     // mark task as waiting not on a barrier
     taskdata->td_taskwait_counter += 1;
     taskdata->td_taskwait_ident = loc;
     taskdata->td_taskwait_thread = gtid + 1;
 #if USE_ITT_BUILD
     // For ITT the taskgroup wait is similar to taskwait until we need to
     // distinguish them
     void *itt_sync_obj = __kmp_itt_taskwait_object(gtid);
-    if (itt_sync_obj != NULL)
+    if (UNLIKELY(itt_sync_obj != NULL))
       __kmp_itt_taskwait_starting(gtid, itt_sync_obj);
 #endif /* USE_ITT_BUILD */
 
 #if OMPT_SUPPORT && OMPT_OPTIONAL
     if (UNLIKELY(ompt_enabled.ompt_callback_sync_region_wait)) {
       ompt_callbacks.ompt_callback(ompt_callback_sync_region_wait)(
           ompt_sync_region_taskgroup, ompt_scope_begin, &(my_parallel_data),
           &(my_task_data), codeptr);
#if OMPT_SUPPORT && OMPT_OPTIONAL
     if (UNLIKELY(ompt_enabled.ompt_callback_sync_region_wait)) {
       ompt_callbacks.ompt_callback(ompt_callback_sync_region_wait)(
           ompt_sync_region_taskgroup, ompt_scope_end, &(my_parallel_data),
           &(my_task_data), codeptr);
     }
 #endif
 
 #if USE_ITT_BUILD
-    if (itt_sync_obj != NULL)
+    if (UNLIKELY(itt_sync_obj != NULL))
       __kmp_itt_taskwait_finished(gtid, itt_sync_obj);
     KMP_FSYNC_ACQUIRED(taskdata); // acquire self - sync with descendants
 #endif /* USE_ITT_BUILD */
   }
   KMP_DEBUG_ASSERT(taskgroup->count == 0);
 
   if (taskgroup->reduce_data != NULL) { // need to reduce?
     int cnt;