There is already some discussion about this topic at the other topic
However, if this is causing you issues that need fixing now, it shouldn't be too hard to work around it manually until a better solution is available
Simplest solution might be to have a script / service running (as root) that does
- read temp from /sys/devices/virtual/thermal/thermal_zone0/temp
- if temp is too high for you, throttle max frequency, e.g. using
cpufreq-set -u 700000
then up the limit again once temp is reduced.
Not ideal of course, but should work in