Real-time scheduling offers priorities 1 (low) to 99 (high). Non-realtime has default of 0, thus lower than lowest RT priority.

Nice value applies to only non-RT processes. The larger the nice value is, the nicer a process is. Am I correct so far?

I'm mainly interested in RT processes now. I see online discussion talking about use of /proc/PID/stat to get priority value, which has a range of -2 to -100 for RT.

This is what confused me - how do we map -2 to -100 into 1 to 99? Is -100 considered the lowest priority thus would map to 1, and -2 would map to 99 [essentially adding 101 to the negative range and we'd get to the proper +ve range]?

Finally, does anyone else find this confusing? I mean, would it be more straightforward if Linux defines say two classes - RT and non-RT, and have 1 single range to describe all priorities?

A web search for "linux nice" turns up many tutorials.

This might be a good one to start with: https://www.lifewire.com/uses-of-com...renice-2201087

To respond to your last question, I don't find it at all confusing because I have no need to use the nice command.

Hey frank bell, I'm trying to understand RT priority mapping, not the nice, or renice. Not at this time anyway.

In particular, I'm trying to map the value returned from reading /proc/PID/stat, which has a range -2 to -100 into a range 1 to 99. It's not very clear how the mapping is done, esp. consider how some online articles use high/low as an adjective to describe nicety.