Real-Time Physics Simulation Forum

I think your problem may be that your not creating the individual solvers in the pool.

This is what I use, and I am stable: auto createSolver = [&]() -> btConstraintSolverPoolMt* { SolverType tmpSolverType = mSolverType; if (tmpSolverType == SolverType::SOLVER_TYPE_SEQUENTIAL_IMPULSE_MT) { // pool solvers shouldn't be parallel solvers, we don't allow that kind of // nested paralle...

So to get contact test to multithread, it's pretty easy. All you need is to add one mutex to collisiondispatcher and then lock it in new/release/dispatch.

infinite performance gains

So id like to have a bunch of btghostobjects and it to me it would make sense to multithread contactTest with the ghost object. I think i just need to add mutexes to various places, anyone have any suggestions? Right now, if i try to call contact test in parallel, it breaks pretty hard.

So i tried handling the collisions in a specific order, assuming they can come in mixed-up, but that does not seem to help. To generate the order, i always attach a unique ID to userIndex when an object is created. obj->setUserIndex(uniqueID++) here's my collision routine. static std::vector< btPers...

Is deterministic MT not supported in bullet 2?

Hello, I am trying to run bullet and use MT, is there a way to do it deterministically? Right now I have the following stuff: setTaskScheduler(SchedulerType::DEFAULT); btDefaultCollisionConstructionInfo cci; cci.m_defaultMaxPersistentManifoldPoolSize = 80000; cci.m_defaultMaxCollisionAlgorithmPoolSi...

OK sorry it looks like the problem was bullet wasn't built with multithreading checked, lol. Sorry!

The following image contains the error and stack trace: https://pasteboard.co/IS8iQqf.png I am using multithreaded bullet, a height map, and a lot of objects. Only happens once objects have started colliding on the heightmap, and only when there's pileup or a lot of interaction. Seems like something...

~btMLCPSolver should delete m_solver; given that m_solver was newed. Right now it leaks instead, especially during: ~btConstraintSolverPoolMt. Which is how I Discovered this issue. Another solution is to not have the ~btConstraintSolverPoolMt delete anything, which I would prefer.

Thanks for helping me. Ok so here was the question from my earlier post explained better I hope: I am using a collision world, not dynamics. I call, Physics::CollisionCallback callback = std::bind(&Playground::collisionCallback, this, std::placeholders::_1, std::placeholders::_2); mPhysics.setCo...

So I have implemented a btGhostObject in my collision world, but It seems to be getting sent to the general collision detection handler. When you said it doesn't collide, were you referring to in a dynamics world? Or am I perhaps doing something wrong?

So my game is top down and sort of rts-ish. A unit has a vision in which it decides to engage an enemy unit. I thought that doing a bullet-collision-shape for sight made sense to gather a list of enemies in range. Is this true?

I want to do my unit sight detection with the library, I think. So, I was thinking to have a collision object, the units sight area, and then if I could just do collision against that object to build a list of sighted objects, is this possible, or is there a better method?

I have just started using bullet. Is this thread related to the master's multithreaded examples?