When we try to test the evk board under the zero temperature. we find the value of temperature always above zero.the code get_temp as follows.
qoriq_thermal.c
static int tmu_get_temp(void *p, int *temp)
{
u32 val;
struct qoriq_tmu_data *data = p;
val = tmu_read(data, &data->regs->site[data->sensor_id].tritsr);
*temp = (val & 0xff) * 1000;
return 0;
}
so when the temperature of cpu under zero, the obtain value will more then the critical value ，lead to system too hot to reboot.
After found this , I print the "val". when the value is above zero,it like this [80000042] and when the value was zero it will be [80000000] or [0]. under the zero will be [70]. then modify the code like this
static int tmu_get_temp(void *p, int *temp)
{
u32 val;
struct qoriq_tmu_data *data = p;
val = tmu_read(data, &data->regs->site[data->sensor_id].tritsr);
*temp = (val & 0xff) * 1000;
/* when the sensor temperature was 0x80000042, we think of it hight the zero,
if the temperature was 0x00000054, we think of it low the zero*/
if(!(val&(1<<31))){
*temp = (-1)*(*temp);
}
return 0;
}
Unfortunately the val is still reverse like this
-----------------------------------------------------------------------------
[ 976.994694] last_temperature=-105000, current_temperature=-105000
[ 979.010619] last_temperature=-105000, current_temperature=-105000
[ 981.026642] last_temperature=-105000, current_temperature=-105000
[ 983.042617] last_temperature=-105000, current_temperature=-105000
[ 985.058700] last_temperature=-105000, current_temperature=105000
[ 985.064762] hantro receive hot notification event: 1
[ 985.069951] System is too hot. GPU3D will work at 1/64 clock.
[ 985.075912] thermal thermal_zone0: critical temperature reached(105 C),shutting down
Message from syslogd@imx8mqevk at Mon Jul 30 10:22:57 2018 ...
imx8mqevk kernel: [ 985.075912] thermal thermal_zone0: critical temperature reached(105 C),shutting down
can you please fix the value of sensor?
