1*4882a593SmuzhiyunToDos sorted by priority: 2*4882a593Smuzhiyun 3*4882a593Smuzhiyun- Use bitmask functions to parse CPU topology more robust 4*4882a593Smuzhiyun (current implementation has issues on AMD) 5*4882a593Smuzhiyun- Try to read out boost states and frequencies on Intel 6*4882a593Smuzhiyun- Somewhere saw the ability to read power consumption of 7*4882a593Smuzhiyun RAM from HW on Intel SandyBridge -> another monitor? 8*4882a593Smuzhiyun- Add another c1e debug idle monitor 9*4882a593Smuzhiyun -> Is by design racy with BIOS, but could be added 10*4882a593Smuzhiyun with a --force option and some "be careful" messages 11*4882a593Smuzhiyun- Add cpu_start()/cpu_stop() callbacks for monitor 12*4882a593Smuzhiyun -> This is to move the per_cpu logic from inside the 13*4882a593Smuzhiyun monitor to outside it. This can be given higher 14*4882a593Smuzhiyun priority in fork_it. 15*4882a593Smuzhiyun- Fork as many processes as there are CPUs in case the 16*4882a593Smuzhiyun per_cpu_schedule flag is set. 17*4882a593Smuzhiyun -> Bind forked process to each cpu. 18*4882a593Smuzhiyun -> Execute start measures via the forked processes on 19*4882a593Smuzhiyun each cpu. 20*4882a593Smuzhiyun -> Run test executable in a forked process. 21*4882a593Smuzhiyun -> Execute stop measures via the forked processes on 22*4882a593Smuzhiyun each cpu. 23*4882a593Smuzhiyun This would be ideal as it will not introduce noise in the 24*4882a593Smuzhiyun tested executable. 25