[ODE] Freezing Problems in ODE.

[Olex]

I'm experiencing same things on both Linux (Gentoo) and WinXP.

I've developed a debug window that shows a percentage of CPU used for
graphics and physics, and everytime during such freezes, physics spike
to 100%.

I'm glad I'm not the only one with this problem. However, I'm not sure
it's ODE's fault. TrueAxis (physics engine, google for it) had the
same problems on Linux as well. The other thing I've noticed is that
on win32, the issue is not always apparent, sometimes after certain
rebuilds of libraries, the issue is not present at all on win32, other
times it does.

It would be helpful to profile the application with ability to analyze
parts of the timeline (during those freeze moments). That should shine
some light on the issue.

On 5/5/06, Matthew Nunya <shatteredmaul at gmail.com> wrote:
>
>
>
> I have been playing around with ODE for a year or so now and have recently
> decided to do something useful with it. I'm currently witting a DLL so that
> I can do some simple physics in "Game Maker" (if you haven't heard of it,
> it's basically a scripting engine of sorts, like Blitz). However, I keep
> running into a nagging problem that won't go away, and I'm sure it isn't
> Game Maker that's causing it. I could be simulating only a couple of bodies
> and joints and everything will be running quite smoothly at 60 fps, then at
> random everything will freeze and my entire system will hang for a few
> seconds before returning back to normal and running smoothly. This could
> happen anytime, it doesn't only occur when things collide.
>
> I have done my best to try and narrow down what's causing it but my attempts
> have been fruitless. I have played with ERP, CFM, step sizes, and other
> configurations for hours to no avail.
>
> What I do know about my problem is that when ever it occurs, the processor
> load spikes at 100% for the duration of the freeze and my hard drive will
> only rev slightly near the end of the freeze. It's very slightly, because
> I'm not running out of physical memory and windows doesn't need to resort to
> virtual memory. In fact, the simulation's memory doesn't change at all. It
> isn't leaking any memory to the best of my knowledge. I am at a real loss as
> to why it happens.
>
> I was wondering if anyone else as experienced this problem and if so, I
> would be incredibly curious to know how to correct it. Any help of advice
> anyone could offer would be infinitely appreciated.
>
> Just in case it helps, this is the system I'm running on:
>
> WinXP (SP2) / Athlon64 3400+ / 1GB Ram / EIDE 160 GB HD
>
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>
>
>


--
Oleksandr Lozitskiy

"A really great talent finds its happiness in execution." by Johann
Wolfgang von Goethe