Skip to contentSkip to conversation

keep ignore_nice_load value when the ondemand govenor is reselected

PATCH In review Eric Piel 1 participant 2 messages Mar 10, 2006 – Mar 23, 2006 CPU frequency scaling framework

lore
Share & tools
lore

Eric Piel

author
#0 lore raw

PATCH 1/1 keep ignore_nice_load value when the ondemand govenor is reselected

Hello,

With the ondemand governor, ignore_nice_load is reset each time the governor is selected. That's annoying because each time I change to another governor and change back to ondemand, I also have to set up ignore_nice_load. eg:

# cat /sys/devices/system/cpu/cpu0/cpufreq/ondemand/ignore_nice_load
1
# cpufreq-set -g powersave
# cpufreq-set -g ondemand
# cat /sys/devices/system/cpu/cpu0/cpufreq/ondemand/ignore_nice_load
0

This behavior is also inconsistent with the other variables of /sys/devices/system/cpu/cpu0/cpufreq/ondemand/

Attached is a patch that corrects the behavior.

see you,
Eric

---
From: Eric Piel <[email protected]>

Keep the value of ignore_nice_load of the ondemand governor even after the governor has been deselected and selected back. This is the behavior of the other exported values of the ondemand governor and it's much more user-friendly.

drivers/cpufreq/cpufreq_ondemand.c +1 -2
@@ -84,6 +84,7 @@ struct dbs_tuners {
 static struct dbs_tuners dbs_tuners_ins = {
 	.up_threshold 		= DEF_FREQUENCY_UP_THRESHOLD,
 	.sampling_down_factor 	= DEF_SAMPLING_DOWN_FACTOR,
+	.ignore_nice		= 0,
 };
 
 static inline unsigned int get_cpu_idle_time(unsigned int cpu)
@@ -431,8 +432,6 @@ static int cpufreq_governor_dbs(struct c
 				def_sampling_rate = MIN_STAT_SAMPLING_RATE;
 
 			dbs_tuners_ins.sampling_rate = def_sampling_rate;
-			dbs_tuners_ins.ignore_nice = 0;
-
 			dbs_timer_init();
 		}
 		

_______________________________________________
Cpufreq mailing list
[email protected]
http://lists.linux.org.uk/mailman/listinfo/cpufreq

author
#1 lore raw

10.03.2006 10:35, Eric Piel wrote/a écrit:

Quoted 10 lines

Hello,

With the ondemand governor, ignore_nice_load is reset each time the governor is selected. That's annoying because each time I change to another governor and change back to ondemand, I also have to set up ignore_nice_load. eg:

# cat /sys/devices/system/cpu/cpu0/cpufreq/ondemand/ignore_nice_load
1
# cpufreq-set -g powersave
# cpufreq-set -g ondemand
# cat /sys/devices/system/cpu/cpu0/cpufreq/ondemand/ignore_nice_load
0
Quoted 3 lines

This behavior is also inconsistent with the other variables of /sys/devices/system/cpu/cpu0/cpufreq/ondemand/

Attached is a patch that corrects the behavior.

Hello, I've sent this patch and "Warn if ondemand governor can not run due to too long transition latency" mostly 2 weeks ago and didn't get any feedback. It would be nice if at least this patch could make it to 2.6.17 as it fixes a glitch. Venki, could you ACK or NACK these patches?

Dave, if they don't apply cleanly against the cpufreq git tree (due to the big clean up commit), I can redo them... just let me know :-)

Eric