aboutsummaryrefslogtreecommitdiff
path: root/rootdir/etc/init.qcom.power.rc
diff options
context:
space:
mode:
authorCosme Domínguez Díaz <cosme.ddiaz@gmail.com>2018-08-09 18:10:47 +0200
committerCosme Domínguez Díaz <cosme.ddiaz@gmail.com>2018-08-09 19:36:45 +0200
commit1e8ebc090ff0aa0334c71f506d029a77f7447a0a (patch)
tree202b736739918f1f0a1f07999014ac21237f41c5 /rootdir/etc/init.qcom.power.rc
parent427d844b840d6d1a1d11ba154c0dbdca5379080b (diff)
msm8996-common: Avoid jumping to the maximum frequency without needing it
* I just don't understand why we have 16 frequencies (from 307 to 1593MHz) on our power cluster and libqti-perfd just wants to force the highest every time I touch the screen... Why implement cpu frequency scaling and then use only two?
Diffstat (limited to 'rootdir/etc/init.qcom.power.rc')
-rw-r--r--rootdir/etc/init.qcom.power.rc3
1 files changed, 3 insertions, 0 deletions
diff --git a/rootdir/etc/init.qcom.power.rc b/rootdir/etc/init.qcom.power.rc
index e1ab531..473bfa3 100644
--- a/rootdir/etc/init.qcom.power.rc
+++ b/rootdir/etc/init.qcom.power.rc
@@ -35,6 +35,9 @@ on init
# Drop write bit from cpu_max_freq so that libqti-perfd can't set it
chmod 0444 /sys/module/msm_performance/parameters/cpu_max_freq
+ # Same with cpu_min_freq to avoid burn our battery
+ chmod 0444 /sys/module/msm_performance/parameters/cpu_min_freq
+
on boot
# Update foreground and background cpusets
write /dev/cpuset/foreground/cpus 0-3