On 06-01-20, 16:30, chenqiwu wrote: > Hhh..I don't agree this, since the cpufreq policy of cpu0 may have > been released before such UAF issue happenning. That won't happen if you do cpufreq_cpu_get(), isn't it ? > By the way, why not get invaild cpufreq policy of another cpu but > only cpu0 here? Probably because this platform has a single cpufreq policy which covers all the CPUs and so it doesn't really matter which CPU you use to get the policy. -- viresh