On 2024-01-18 08:38:12 [-0800], Jakub Kicinski wrote: > > If this is the per-CPU BH lock (which I want to remove) then it will > > continue until all softirqs complete. > > So there's no way for a process to know on RT that someone with higher > prio is waiting for it to release its locks? :( You could add a function to check if your current priority is inherited from someone else and if so start dropping the locks you think are responsible for it. I made a PoC that appears to work for timer_list timer which is one of the softirqs. This made me realise that I need in more spots and I am doing it for the wrong reasons… Sebastian
