Skip to contentSkip to conversation

sched: Fix boolean expression in move_tasks()

PATCH In review Peter Williams 2 participants 4 messages Apr 24, 2006 – Apr 25, 2006

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 sched: Fix boolean expression in move_tasks()

Problem:

In the patch sched-avoid-unnecessarily-moving-highest-priority-task-move_tasks.patch
I got a the sense of a boolean expression wrong when assigning a value to skip_for_load. The expression should have been negated before being assigned.

Solution:

Negate the expression and apply de Marcos rule to simplify it. This patch is on top of sched-avoid-unnecessarily-moving-highest-priority-task-move_tasks.patch

Signature · 4 lines
Peter Williams                                   [email protected]

"Learning, n. The kind of ignorance distinguishing the studious."
 -- Ambrose Bierce
kernel/sched.c +1 -1
Index: MM-2.6.17-rc1-mm3/kernel/sched.c
===================================================================
@@ -2108,7 +2108,7 @@ skip_queue:
 	 */
 	skip_for_load = tmp->load_weight > rem_load_move;
 	if (skip_for_load && idx < this_best_prio)
-		skip_for_load = busiest_best_prio_seen || idx != busiest_best_prio;
+		skip_for_load = !busiest_best_prio_seen && idx == busiest_best_prio;
 	if (skip_for_load ||
 	    !can_migrate_task(tmp, busiest, this_cpu, sd, idle, &pinned)) {
 		if (curr != head)
#1 lore raw
Quoted 14 lines · Peter Williams

Negate the expression and apply de Marcos rule to simplify it. This
patch is on top of sched-avoid-unnecessarily-moving-highest-priority-task-move_tasks.patch

--
Peter Williams [email protected]

"Learning, n. The kind of ignorance distinguishing the studious."
-- Ambrose Bierce

Quoted 11 lines

[smpnice-fix-boolean-expression text/plain (824 bytes)]
Index: MM-2.6.17-rc1-mm3/kernel/sched.c
===================================================================
--- MM-2.6.17-rc1-mm3.orig/kernel/sched.c 2006-04-21 12:26:54.000000000 +1000
+++ MM-2.6.17-rc1-mm3/kernel/sched.c 2006-04-25 09:09:54.000000000 +1000
@@ -2108,7 +2108,7 @@ skip_queue:
*/
skip_for_load = tmp->load_weight > rem_load_move; if (skip_for_load && idx < this_best_prio)
- skip_for_load = busiest_best_prio_seen || idx != busiest_best_prio;
+ skip_for_load = !busiest_best_prio_seen && idx == busiest_best_prio;

But Suresh's sched-avoid-unnecessarily-moving-highest-priority-task-move_tasks-fix.patch changed all this code:

	/*
	 * To help distribute high priority tasks accross CPUs we don't
	 * skip a task if it will be the highest priority task (i.e. smallest
	 * prio value) on its new queue regardless of its load weight
	 */
	skip_for_load = tmp->load_weight > rem_load_move;
	if (skip_for_load && idx < this_best_prio && idx == busiest_best_prio)
		skip_for_load = !busiest_best_prio_seen &&
				head->next == head->prev;
	if (skip_for_load ||
	    !can_migrate_task(tmp, busiest, this_cpu, sd, idle, &pinned)) {
		if (curr != head)
			goto skip_queue;
		idx++;
		goto skip_bitmap;
	}

What to do?

#2 lore raw
Quoted 48 lines · Andrew Morton
Quoted 13 lines · Peter Williams

Negate the expression and apply de Marcos rule to simplify it. This
patch is on top of sched-avoid-unnecessarily-moving-highest-priority-task-move_tasks.patch

--
Peter Williams [email protected]

"Learning, n. The kind of ignorance distinguishing the studious."
-- Ambrose Bierce

Quoted 11 lines

[smpnice-fix-boolean-expression text/plain (824 bytes)]
Index: MM-2.6.17-rc1-mm3/kernel/sched.c
===================================================================
--- MM-2.6.17-rc1-mm3.orig/kernel/sched.c 2006-04-21 12:26:54.000000000 +1000
+++ MM-2.6.17-rc1-mm3/kernel/sched.c 2006-04-25 09:09:54.000000000 +1000
@@ -2108,7 +2108,7 @@ skip_queue:
*/
skip_for_load = tmp->load_weight > rem_load_move; if (skip_for_load && idx < this_best_prio)
- skip_for_load = busiest_best_prio_seen || idx != busiest_best_prio;
+ skip_for_load = !busiest_best_prio_seen && idx == busiest_best_prio;

But Suresh's sched-avoid-unnecessarily-moving-highest-priority-task-move_tasks-fix.patch changed all this code:

	/*
	 * To help distribute high priority tasks accross CPUs we don't
	 * skip a task if it will be the highest priority task (i.e. smallest
	 * prio value) on its new queue regardless of its load weight
	 */
	skip_for_load = tmp->load_weight > rem_load_move;
	if (skip_for_load && idx < this_best_prio && idx == busiest_best_prio)
		skip_for_load = !busiest_best_prio_seen &&
				head->next == head->prev;
	if (skip_for_load ||
	    !can_migrate_task(tmp, busiest, this_cpu, sd, idle, &pinned)) {
		if (curr != head)
			goto skip_queue;
		idx++;
		goto skip_bitmap;
	}
Quoted 1 line

What to do?

Suresh's patch was wrong and this was intended as an alternative. Unfortunately, it is also in adequate and the setting of busiest_best_prio_seen needs to be moved to just after skip_for_load is set. I have another patch that does that (and adds to the comments). Should I send that separately or roll the two patches together?

Peter

Signature · 4 lines
Peter Williams                                   [email protected]

"Learning, n. The kind of ignorance distinguishing the studious."
 -- Ambrose Bierce
#3 lore raw
Quoted 4 lines · Peter Williams
Quoted 2 lines

What to do?

Suresh's patch was wrong and this was intended as an alternative.

OIC.

Quoted 4 lines

Unfortunately, it is also in adequate and the setting of busiest_best_prio_seen needs to be moved to just after skip_for_load is set. I have another patch that does that (and adds to the comments). Should I send that separately or roll the two patches together?

I don't mind, really - rolled together, I guess.