MINOR: tasks: only visit filled task slots after processing them
process_runnable_tasks() needs to requeue or wake up tasks after
processing them in batches. By only refilling the existing ones, we
avoid revisiting all the queue. The performance gain is measurable
starting with two threads, where the request rate climbs to 657k/s
compared to 644k.
diff --git a/src/task.c b/src/task.c
index 37994a3..3d61f98 100644
--- a/src/task.c
+++ b/src/task.c
@@ -188,6 +188,7 @@
struct eb32sc_node *rq_next;
struct task *local_tasks[16];
int local_tasks_count;
+ int final_tasks_count;
tasks_run_queue_cur = tasks_run_queue; /* keep a copy for reporting */
nb_tasks_cur = nb_tasks;
max_processed = tasks_run_queue;
@@ -241,6 +242,7 @@
SPIN_UNLOCK(TASK_RQ_LOCK, &rq_lock);
+ final_tasks_count = 0;
for (i = 0; i < local_tasks_count ; i++) {
t = local_tasks[i];
/* This is an optimisation to help the processor's branch
@@ -250,24 +252,23 @@
t = process_stream(t);
else
t = t->process(t);
- local_tasks[i] = t;
+ if (t)
+ local_tasks[final_tasks_count++] = t;
}
SPIN_LOCK(TASK_RQ_LOCK, &rq_lock);
- for (i = 0; i < local_tasks_count ; i++) {
+ for (i = 0; i < final_tasks_count ; i++) {
t = local_tasks[i];
- if (likely(t != NULL)) {
- t->state &= ~TASK_RUNNING;
- /* If there is a pending state
- * we have to wake up the task
- * immediatly, else we defer
- * it into wait queue
- */
- if (t->pending_state)
- __task_wakeup(t);
- else
- task_queue(t);
- }
+ t->state &= ~TASK_RUNNING;
+ /* If there is a pending state
+ * we have to wake up the task
+ * immediatly, else we defer
+ * it into wait queue
+ */
+ if (t->pending_state)
+ __task_wakeup(t);
+ else
+ task_queue(t);
}
} while (max_processed > 0);