On 06.06.2011 18:36, Roland Häder wrote: > Wouldn't it be possible to set lower priority for those loader > processes? I think no user can be asked to buy an expensive SCSI > hardware system (which would lower the CPU usage compared to IDE/SATA > systems). > > I mean somthing like "niceness" for forked loader threads, like we have > on Uni*/Linux boxes. So the "main" simulation thread would have more CPU > cycles left than the loader (which can quietly load in background).
You're (and everyone is) welcome to investigate the details of these hickups in more detail. That would be a huge improvement for FG to get rid/lessen these issues. However, I'm pretty sure it's not as simple as changing a thread priority. These hickups also happen on multi-core CPUs. I can see them on my local 6-core machine - and we saw them on our dual CPU/12-core monster at LinuxTag - where each thread basically had its own CPU core (so priority doesn't matter much). My guess is that there is an issue blocking our main loop when new models are loaded. Possibly the mainloop is blocked in some mutex while a new model is merged into the scene graph. But it'd be up to someone to investigate this in more detail (and hopefully also find a solution). cheers, Thorsten ------------------------------------------------------------------------------ Simplify data backup and recovery for your virtual environment with vRanger. Installation's a snap, and flexible recovery options mean your data is safe, secure and there when you need it. Discover what all the cheering's about. Get your free trial download today. http://p.sf.net/sfu/quest-dev2dev2 _______________________________________________ Flightgear-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/flightgear-devel
