[MEDIUM] introduce task->nice and boot access to statistics

The run queue scheduler now considers task->nice to queue a task and
to pick a task out of the queue. This makes it possible to boost the
access to statistics (both via HTTP and UNIX socket). The UNIX socket
receives twice as much a boost as the HTTP socket because it is more
sensible.

include/proto/task.h

@@ -32,7 +32,8 @@
 
 #include <types/task.h>
 
-extern unsigned int run_queue;  /* run queue size */
+extern unsigned int run_queue;    /* run queue size */
+extern unsigned int niced_tasks;  /* number of niced tasks in the run queue */
 extern struct pool_head *pool2_task;
 
 /* perform minimal initializations, report 0 in case of error, 1 if OK. */
@@ -50,6 +51,8 @@ static inline struct task *task_sleep(struct task *t)
 		t->state = TASK_IDLE;
 		eb32_delete(&t->eb);
 		run_queue--;
+		if (likely(t->nice))
+			niced_tasks--;
 	}
 	return t;
 }
@@ -65,9 +68,11 @@ static inline struct task *task_delete(struct task *t)
 	if (t->eb.node.leaf_p)
 		eb32_delete(&t->eb);
 
-	if (t->state == TASK_RUNNING)
+	if (t->state == TASK_RUNNING) {
 		run_queue--;
-
+		if (likely(t->nice))
+			niced_tasks--;
+	}
 	return t;
 }
 
@@ -79,6 +84,7 @@ static inline struct task *task_init(struct task *t)
 {
 	t->eb.node.leaf_p = NULL;
 	t->state = TASK_IDLE;
+	t->nice = 0;
 	return t;
 }
 

include/types/task.h

@@ -39,6 +39,7 @@ struct task {
 	struct timeval expire;		/* next expiration time for this task, use only for fast sorting */
 	void (*process)(struct task *t, struct timeval *next);	/* the function which processes the task */
 	void *context;			/* the task's context */
+	int nice;			/* the task's current nice value from -1024 to +1024 */
 };
 
 #endif /* _TYPES_TASK_H */

src/proto_http.c

@@ -5160,6 +5160,7 @@ int stats_check_uri_auth(struct session *t, struct proxy *backend)
 	t->logs.tv_request = now;
 	t->data_source = DATA_SRC_STATS;
 	t->data_state  = DATA_ST_INIT;
+	t->task->nice = -32; /* small boost for HTTP statistics */
 	produce_content(t);
 	return 1;
 }

src/proto_uxst.c

@@ -447,6 +447,7 @@ int uxst_event_accept(int fd) {
 		task_init(t);
 		t->process = l->handler;
 		t->context = s;
+		t->nice = -64;  /* we want to boost priority for local stats */
 
 		s->task = t;
 		s->fe = NULL;

src/task.c

@@ -24,6 +24,7 @@
 struct pool_head *pool2_task;
 
 unsigned int run_queue = 0;
+unsigned int niced_tasks = 0; /* number of niced tasks in the run queue */
 
 /* Principle of the wait queue.
  *
@@ -125,7 +126,11 @@ int init_task()
 	return pool2_task != NULL;
 }
 
-/* puts the task <t> in run queue <q>, and returns <t> */
+/* Puts the task <t> in run queue at a position depending on t->nice.
+ * <t> is returned. The nice value assigns boosts in 32th of the run queue
+ * size. A nice value of -1024 sets the task to -run_queue*32, while a nice
+ * value of 1024 sets the task to run_queue*32.
+ */
 struct task *task_wakeup(struct task *t)
 {
 	if (t->state == TASK_RUNNING)
@@ -136,6 +141,18 @@ struct task *task_wakeup(struct task *t)
 
 	run_queue++;
 	t->eb.key = ++rqueue_ticks;
+
+	if (likely(t->nice)) {
+		int offset;
+
+		niced_tasks++;
+		if (likely(t->nice > 0))
+			offset = (unsigned)((run_queue * (unsigned int)t->nice) / 32U);
+		else
+			offset = -(unsigned)((run_queue * (unsigned int)-t->nice) / 32U);
+		t->eb.key += offset;
+	}
+
 	t->state  = TASK_RUNNING;
 
 	eb32_insert(&rqueue[ticks_to_tree(t->eb.key)], &t->eb);
@@ -218,10 +235,10 @@ void wake_expired_tasks(struct timeval *next)
  * used to assign a position to each task. This counter may be combined with
  * other variables (eg: nice value) to set the final position in the tree. The
  * counter may wrap without a problem, of course. We then limit the number of
- * tasks processed at once to 1/8 of the number of tasks in the queue, so that
- * general latency remains low and so that task positions have a chance to be
- * considered. It also reduces the number of trees to be evaluated when no task
- * remains.
+ * tasks processed at once to 1/4 of the number of tasks in the queue, and to
+ * 200 max in any case, so that general latency remains low and so that task
+ * positions have a chance to be considered. It also reduces the number of
+ * trees to be evaluated when no task remains.
  *
  * Just like with timers, we start with tree[(current - 1)], which holds past
  * values, and stop when we reach the middle of the list. In practise, we visit
@@ -240,7 +257,12 @@ void process_runnable_tasks(struct timeval *next)
 	if (!run_queue)
 		return;
 
-	max_processed = (run_queue + 7) / 8;
+	max_processed = run_queue;
+	if (max_processed > 200)
+		max_processed = 200;
+
+	if (likely(niced_tasks))
+		max_processed /= 4;
 
 	tree = ticks_to_tree(rqueue_ticks);
 	stop = (tree + TIMER_TREES / 2) & TIMER_TREE_MASK;
@@ -255,6 +277,8 @@ void process_runnable_tasks(struct timeval *next)
 			eb = eb32_next(eb);
 
 			run_queue--;
+			if (likely(t->nice))
+				niced_tasks--;
 			t->state = TASK_IDLE;
 			eb32_delete(&t->eb);