[Bloat] Bad shaping at low rates, after commit 56b765b79 (htb: improved accuracy at high rates)

[Eric Dumazet]

On Tue, 2013-06-04 at 08:18 -0700, Eric Dumazet wrote:

> I have a good idea of what's going on for htb at low rates, I am testing
> a fix, thanks for the report !
> 

Yes, we need to convert whole thing to use ns units, instead
of a mix of 64ns and 1ns units.

Please test the following patch :

diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c
index 79b1876..6c53341 100644
--- a/net/sched/sch_htb.c
+++ b/net/sched/sch_htb.c
@@ -109,7 +109,7 @@ struct htb_class {
 	} un;
 	struct rb_node node[TC_HTB_NUMPRIO];	/* node for self or feed tree */
 	struct rb_node pq_node;	/* node for event queue */
-	psched_time_t pq_key;
+	s64 pq_key;
 
 	int prio_activity;	/* for which prios are we active */
 	enum htb_cmode cmode;	/* current mode of the class */
@@ -124,7 +124,7 @@ struct htb_class {
 	s64 buffer, cbuffer;	/* token bucket depth/rate */
 	psched_tdiff_t mbuffer;	/* max wait time */
 	s64 tokens, ctokens;	/* current number of tokens */
-	psched_time_t t_c;	/* checkpoint time */
+	s64 t_c;		/* checkpoint time */
 };
 
 struct htb_sched {
@@ -141,7 +141,7 @@ struct htb_sched {
 	struct rb_root wait_pq[TC_HTB_MAXDEPTH];
 
 	/* time of nearest event per level (row) */
-	psched_time_t near_ev_cache[TC_HTB_MAXDEPTH];
+	s64 near_ev_cache[TC_HTB_MAXDEPTH];
 
 	int defcls;		/* class where unclassified flows go to */
 
@@ -149,7 +149,7 @@ struct htb_sched {
 	struct tcf_proto *filter_list;
 
 	int rate2quantum;	/* quant = rate / rate2quantum */
-	psched_time_t now;	/* cached dequeue time */
+	s64 now;	/* cached dequeue time */
 	struct qdisc_watchdog watchdog;
 
 	/* non shaped skbs; let them go directly thru */
@@ -664,8 +664,8 @@ static void htb_charge_class(struct htb_sched *q, struct htb_class *cl,
  * next pending event (0 for no event in pq, q->now for too many events).
  * Note: Applied are events whose have cl->pq_key <= q->now.
  */
-static psched_time_t htb_do_events(struct htb_sched *q, int level,
-				   unsigned long start)
+static s64 htb_do_events(struct htb_sched *q, int level,
+			 unsigned long start)
 {
 	/* don't run for longer than 2 jiffies; 2 is used instead of
 	 * 1 to simplify things when jiffy is going to be incremented
@@ -857,7 +857,7 @@ static struct sk_buff *htb_dequeue(struct Qdisc *sch)
 	struct sk_buff *skb;
 	struct htb_sched *q = qdisc_priv(sch);
 	int level;
-	psched_time_t next_event;
+	s64 next_event;
 	unsigned long start_at;
 
 	/* try to dequeue direct packets as high prio (!) to minimize cpu work */
@@ -880,7 +880,7 @@ ok:
 	for (level = 0; level < TC_HTB_MAXDEPTH; level++) {
 		/* common case optimization - skip event handler quickly */
 		int m;
-		psched_time_t event;
+		s64 event;
 
 		if (q->now >= q->near_ev_cache[level]) {
 			event = htb_do_events(q, level, start_at);
@@ -1200,7 +1200,7 @@ static void htb_parent_to_leaf(struct htb_sched *q, struct htb_class *cl,
 	parent->un.leaf.q = new_q ? new_q : &noop_qdisc;
 	parent->tokens = parent->buffer;
 	parent->ctokens = parent->cbuffer;
-	parent->t_c = psched_get_time();
+	parent->t_c = ktime_to_ns(ktime_get());
 	parent->cmode = HTB_CAN_SEND;
 }
 
@@ -1417,8 +1417,8 @@ static int htb_change_class(struct Qdisc *sch, u32 classid,
 		/* set class to be in HTB_CAN_SEND state */
 		cl->tokens = PSCHED_TICKS2NS(hopt->buffer);
 		cl->ctokens = PSCHED_TICKS2NS(hopt->cbuffer);
-		cl->mbuffer = 60 * PSCHED_TICKS_PER_SEC;	/* 1min */
-		cl->t_c = psched_get_time();
+		cl->mbuffer = 60ULL * NSEC_PER_SEC;	/* 1min */
+		cl->t_c = ktime_to_ns(ktime_get());
 		cl->cmode = HTB_CAN_SEND;
 
 		/* attach to the hash list and parent's family */