On Tue, 8 Jul 2008, Daniel Gollub wrote: > > > >> Got a couple of requests for mlockall() in cyclictest, so I added an > > > >> --mlockall > > > >> option to cyclictest, signaltest and pi_stress. > > > > > > > > One small remark: it would be better to do the mlockall() by default. > > > > As far as I know this should be done always for real-time applications. > > > > I am actually surprised that it is not done already. It is also > > > > mentioned in the Howto build an RT-application on the RT linux wiki. > > > > > > > > Robert > > > > > > I thought about that, but figured it would be better just to have the > > > option rather than change long-standing behavior. Right. Nope. We dont want to change the applications default behaviour. Also mlockall is only relevant if your system supports swap. > > > > > > It's Thomas's call though :) > > > > I prefer an option. That gives the mm folks less of an excuse why they > > swap out hot memory :) Hehe > Hi Thomas, > > unfortunately this nice feature didn't made it into rt-tests 0.23 release. Is > there any plan to include this for upcoming rt-tests releases? Yup. Thanks for the reminder. It just got lost in my ever growing inbox :( Thanks, tglx -- To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
