lkml.org 
[lkml]   [2002]   [Oct]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[rfc][patch] kernel/sched.c oddness?
This snippet of code appears wrong...  Either that, or the accompanying 
comment is wrong?

from kernel/sched.c::find_busiest_queue():

|
*imbalance = (max_load - nr_running) / 2;
|
|
/* It needs an at least ~25% imbalance to trigger balancing. */
|
if (!idle && (*imbalance < (max_load + 3)/4)) {
|
busiest = NULL;
|
goto out;
|
}

The comment says 25% imbalance, but the code is really checking for a
~50% imbalance. The attatched patch moves the division by two to the
pull_task function where the imbalance number is actually used. This
patch makes the code match the comment, and divides the imbalance by two
where it is needed.

Please let me know if I've misinterpreted what this code is supposed to
be doing, -or- if we really want the comment to match the current code.

Cheers!

-Matt
diff -Nur --exclude-from=/usr/src/.dontdiff linux-2.5.40-vanilla/kernel/sched.c linux-2.5.40-sched_cleanup/kernel/sched.c
--- linux-2.5.40-vanilla/kernel/sched.c Tue Oct 1 00:07:35 2002
+++ linux-2.5.40-sched_cleanup/kernel/sched.c Wed Oct 2 11:22:41 2002
@@ -689,7 +689,7 @@
if (likely(!busiest))
goto out;

- *imbalance = (max_load - nr_running) / 2;
+ *imbalance = (max_load - nr_running);

/* It needs an at least ~25% imbalance to trigger balancing. */
if (!idle && (*imbalance < (max_load + 3)/4)) {
@@ -746,6 +746,11 @@
task_t *tmp;

busiest = find_busiest_queue(this_rq, this_cpu, idle, &imbalance);
+ /*
+ * We only want to steal a number of tasks equal to 1/2 the imbalance,
+ * otherwise, we'll just shift the imbalance to the new queue.
+ */
+ imbalance >>= 1;
if (!busiest)
goto out;
\
 
 \ /
  Last update: 2005-03-22 13:29    [W:0.086 / U:1.648 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site