[PATCH]: Check for leaf-class when classifying by priority

[PATCH]: Check for leaf-class when classifying by priority

[Patrick McHardy]

[-- Attachment #1: Type: text/plain, Size: 224 bytes --]

Hi devik,

this patch makes htb_classify check if a class is a leaf when classifying
by priority. It also adds a check from cbq to skip htb_find if the majors
of skb->priority and sch->handle differ.

Best regards,
Patrick


[-- Attachment #2: htb-classify.diff --]
[-- Type: text/plain, Size: 1145 bytes --]

# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher.
# This patch includes the following deltas:
#	           ChangeSet	1.1430  -> 1.1431 
#	 net/sched/sch_htb.c	1.15    -> 1.16   
#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 03/11/12	kaber@trash.net	1.1431
# Check for leaf-class when classifying by priority, don't try if majors differ
# --------------------------------------------
#
diff -Nru a/net/sched/sch_htb.c b/net/sched/sch_htb.c
--- a/net/sched/sch_htb.c	Thu Nov 13 15:52:28 2003
+++ b/net/sched/sch_htb.c	Thu Nov 13 15:52:28 2003
@@ -303,8 +303,10 @@
 	   rules in it */
 	if (skb->priority == sch->handle)
 		return HTB_DIRECT;  /* X:0 (direct flow) selected */
-	if ((cl = htb_find(skb->priority,sch)) != NULL) 
-		return cl;
+	if (TC_H_MAJ(skb->priority^sch->handle) == 0 &&
+	    (cl = htb_find(skb->priority,sch)) != NULL) 
+		if (cl->level == 0)
+			return cl;
 
 	tcf = q->filter_list;
 	while (tcf && (result = tc_classify(skb, tcf, &res)) >= 0) {

Re: [PATCH]: Check for leaf-class when classifying by priority

[David S. Miller]

On Thu, 13 Nov 2003 16:00:50 +0100
Patrick McHardy <kaber@trash.net> wrote:

> this patch makes htb_classify check if a class is a leaf when classifying
> by priority. It also adds a check from cbq to skip htb_find if the majors
> of skb->priority and sch->handle differ.

I'm going to hold on this patch for two reasons:

1) To allow Martin to respond and review it.
2) I want the OOPS'er we have pending on the HTB scheduler
   resolved first.

Martin, have you made any progress with that HTB 2.6.x oops
issue yet?

Thanks guys.