ARM: dts: msm: Update GPU thermal rule for lahaina

GPU thermal mitigation configuration uses a virtual sensor, which will
estimate the max of two GPU sensors. Change the GPU thermal rule to
directly use these individual sensors in two separate rule to achieve a
better thermal control.

Change-Id: I05ab74fd63657cefd6b251cf6528e74adf4e636d
This commit is contained in:
Ram Chandrasekar 2020-08-24 14:29:01 -07:00
commit 2d2cfee8db

View file

@ -974,14 +974,13 @@
};
};
gpuss-max-step {
gpuss-0-step {
polling-delay-passive = <10>;
polling-delay = <100>;
polling-delay = <0>;
thermal-governor = "step_wise";
thermal-sensors = <&tsens1 1>, <&tsens1 2>;
sensor-aggregation = <AGGREGATE_MAX_VALUE>;
thermal-sensors = <&tsens1 1>;
trips {
gpuss_config: active-config0 {
gpuss0_config: active-config0 {
temperature = <95000>;
hysteresis = <0>;
type = "passive";
@ -990,7 +989,29 @@
cooling-maps {
gpu_cdev {
trip = <&gpuss_config>;
trip = <&gpuss0_config>;
cooling-device = <&msm_gpu THERMAL_NO_LIMIT
THERMAL_NO_LIMIT>;
};
};
};
gpuss-1-step {
polling-delay-passive = <10>;
polling-delay = <0>;
thermal-governor = "step_wise";
thermal-sensors = <&tsens1 2>;
trips {
gpuss1_config: active-config0 {
temperature = <95000>;
hysteresis = <0>;
type = "passive";
};
};
cooling-maps {
gpu_cdev {
trip = <&gpuss1_config>;
cooling-device = <&msm_gpu THERMAL_NO_LIMIT
THERMAL_NO_LIMIT>;
};