| Message ID | 20191104133020.8820-6-p.paillet@st.com (mailing list archive) |
|---|---|
| State | New, archived |
| Delegated to: | Eduardo Valentin |
| Headers | show |
| Series | thermal: stm32: driver improvements | expand |
diff --git a/drivers/thermal/st/stm_thermal.c b/drivers/thermal/st/stm_thermal.c index 9986716b17c1..f7168762fbde 100644 --- a/drivers/thermal/st/stm_thermal.c +++ b/drivers/thermal/st/stm_thermal.c @@ -324,7 +324,8 @@ static int stm_thermal_set_trips(void *data, int low, int high) if (low > -INT_MAX) { sensor->low_temp_enabled = 1; - ret = stm_thermal_calculate_threshold(sensor, low, &th); + /* add 0.5 of hysteresis due to measurement error */ + ret = stm_thermal_calculate_threshold(sensor, low - 500, &th); if (ret) return ret;
With the STM32 thermal peripheral, it is not possible to dump the temperature that has caused the interrupt. When the temperature reaches the low threshold, we generally read a temperature that is a little bit higher than the low threshold. This maybe due to sampling precision, and also because the CPU becomes hotter when it quits WFI mode. In that case, the framework does not change the trip points. This leads to a lot of low threshold interrupts. The fix is to set the low threshold value 0.5 degrees Celsius below the actual request. The problem is not so frequent with the high threshold and it would no be a good idea to set the threshold value higher than the request. Signed-off-by: Pascal Paillet <p.paillet@st.com> --- drivers/thermal/st/stm_thermal.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)