Right, indeed the timerslack (50us) is not added if the process is real-time. But the C-state problem can still hit you hard.
Cheers, Vincenzo Il giorno lun 29 ott 2018 alle ore 16:23 Stephen Hemminger < step...@networkplumber.org> ha scritto: > On Sun, 28 Oct 2018 09:07:02 +0100 > Vincenzo Maffione <v.maffi...@gmail.com> wrote: > > > If you don't care about CPU utilization, busy-wait is the way to go. > > Otherwise you can try to usleep() for something like 60-70 us (to lower > the > > CPU utilization), and then busy-wait for the rest of the interval (so > that > > you still have maximum precision). > > > > You may want to have a look at this > > https://academic.oup.com/comjnl/article/61/6/808/4259797 , specially > > sections 4.3 and 4.1. > > > > Cheers, > > Vincenzo > > > > Il giorno dom 28 ott 2018 alle ore 09:01 Sungho Hong < > > maverickji...@gmail.com> ha scritto: > > > > > Thank you very much for the reply. > > > But in that case, should I busy wait, to create the delay? > > > > > > > > > > > > > If doing very brief and exact sleep, you should change your thread > priority to > one of the Real Time classes. > -- Vincenzo