Apply clang-format-all.sh using the _clang-format file through all the cpp/.h files.
make sure not to apply it to certain serialization structures, since some parser expects the * as part of the name, instead of type.
This commit contains no other changes aside from adding and applying clang-format-all.sh
This avoids issues with systems with large mass ratios.
Test: add this to BasicDemo/BasicExample.cpp in initPhysics
m_dynamicsWorld->getSolverInfo().m_numIterations = 1000;
m_dynamicsWorld->getSolverInfo().m_leastSquaresResidualThreshold = 1e-4;
- threading: adding btSequentialImpulseConstraintSolverMt
- task scheduler: added parallelSum so that parallel solver can compute residuals
- CommonRigidBodyMTBase: add slider for solver least squares residual and allow multithreading without needing OpenMP, TBB, or PPL
- taskScheduler: don't wait for workers to sleep/signal at the end of each parallel block
- parallel solver: convertContacts split into an allocContactConstraints and setupContactConstraints stage, the latter of which is done in parallel
- parallel solver: rolling friction is now interleaved along with normal friction
- parallel solver: batchified split impulse solving + some cleanup
- parallel solver: sorting batches from largest to smallest
- parallel solver: added parallel batch creation
- parallel solver: added warmstartingWriteBackContacts func + other cleanup
- task scheduler: truncate low bits to preserve determinism with parallelSum
- parallel solver: reducing dynamic mem allocs and trying to parallelize more of the batch setup
- parallel solver: parallelize updating constraint batch ids for merging
- parallel solver: adding debug visualization
- task scheduler: make TBB task scheduler parallelSum deterministic
- parallel solver: split batch gen code into separate file; allow selection of batch gen method
- task scheduler: add sleepWorkerThreadsHint() at end of simulation
- parallel solver: added grain size per phase
- task Scheduler: fix for strange threading issue; also no need for main thread to wait for workers to sleep
- base constraint solver: break out joint setup into separate function for profiling/overriding
- parallel solver: allow different batching method for contacts vs joints
- base constraint solver: add convertJoint and convertBodies to make it possible to parallelize joint and body conversion
- parallel solver: convert joints and bodies in parallel now
- parallel solver: speed up batch creation with run-length encoding
- parallel solver: batch gen: run-length expansion in parallel; collect constraint info in parallel
- parallel solver: adding spatial grid batching method
- parallel solver: enhancements to spatial grid batching
- sequential solver: moving code for writing back into functions that derived classes can call
- parallel solver: do write back of bodies and joints in parallel
- parallel solver: removed all batching methods except for spatial grid (others were ineffective)
- parallel solver: added 2D or 3D grid batching options; and a bit of cleanup
- move btDefaultTaskScheduler into LinearMath project
Bullet C-API b3ChangeDynamicsInfoSetSpinningFriction/RollingFriction/Resitution
b3PhysicsParamSetRestitutionVelocityThreshold, / pybullet.setPhysicsEngineParameter restitutionVelocityThreshold:
if the velocity is below this threshhold, the restitution is zero (this prevents energy buildup at near-resting state)
pybullet restitution.py example.
implement friction anchors, position friction correction, disabled by default. Use colObj->setCollisionFlag(flag | CF_HAS_FRICTION_ANCHOR); See test/RobotClientAPI/SlopeFrictionMain.cpp. In URDF or SDF, add <friction_anchor/> in <contact> section of <link> to enable.
PhysicsServer: properly restore old activation state after releasing picked object
btMultiBodyConstraintSolver: disable flip/flop of contact/friction constraint solving by default (it breaks some internal flaky unit tests)
(for example solverInfo().m_leastSquaresResidualThreshold = 1e-7 and use large m_numSolverIterations
disable sphere-sphere contact cache, it is buggy (some contact point stay in the cache, when sphere penetrates more than total margins)
tweak some gpu demo settings
- fixing various race conditions throughout (usage of static vars, etc)
- addition of a few lightweight mutexes (which are compiled out by default)
- slight code rearrangement in discreteDynamicsWorld to facilitate multithreading
- PoolAllocator::allocate() can now be called when pool is full without
crashing (null pointer returned)
- PoolAllocator allocate and freeMemory, are OPTIONALLY threadsafe
(default is un-threadsafe)
- CollisionDispatcher no longer checks if the pool allocator is full
before calling allocate(), instead it just calls allocate() and
checks if the return is null -- this avoids a race condition
- SequentialImpulseConstraintSolver OPTIONALLY uses different logic in
getOrInitSolverBody() to avoid a race condition with kinematic bodies
- addition of 2 classes which together allow simulation islands to be run
in parallel:
- btSimulationIslandManagerMt
- btDiscreteDynamicsWorldMt
- MultiThreadedDemo example in the example browser demonstrating use of
OpenMP, Microsoft PPL, and Intel TBB
- use multithreading for other demos
- benchmark demo: add parallel raycasting
contact force problems". The solver needs a lot of iterations, before the quality goes up (~ 1000)
Thanks to Gabor PUHR for the contribution!
Improved the btLemkeSolver.
Remove the sparse optimizations from the btMatrixX.h, replace it with explicit call to rowComputeNonZeroElements (only used in the btSolveProjectedGaussSeidel), it was likely slowing things down, without being useful.
Re-enable SIMD in the solver (was accidently disabled in Bullet 2.82 release)
Apply patch for CMake config, see Issue 754 (Issue 753)
Fix a few issue with the MLCP solver: allow split impulse, and fix offset in friction dependencies
Added a PGS solver that uses the sparsity of the A matrix, just for testing (the equivalent sequential impulse solver is much faster, not having to allocate the big matrices)
The original version was written by Stephen Thompson.
I replaced Eigen math by Bullet LinearMath, and added a dedicated 6x6 matrix solver.
Also I integrated support for collisions/contact constraints between btMultiBody and btRigidBody, and de-activation support.
See Demos/FeatherstoneMultiBodyDemo/Win32FeatherstoneMultiBodyDemo.cpp for example usage.
There is currently only support for contact constraints for btMultiBody.
Next on the list will be adding support for joint limit constraint for btMultiBody.
The implementation is still experimental/untested, the quality will improve in upcoming Bullet releases.
See http://www.youtube.com/watch?v=RV7sBAsKu4M and Bullet/Demos/RollingFrictionDemo
Fixes in FractureDemo (mouse picking constraint needs to be removed, otherwise constraint solver crashes/asserts)
warmstarting for contact points was broken, fix in btPersistentManifold
enable split impulse by default (at the cost of some performance)
add the option for zero-length friction (instead of recomputing friction directions using btPlaneSpace), use the solver mode flag SOLVER_ALLOW_ZERO_LENGTH_FRICTION_DIRECTIONS
precompute lateral friction directions (in btManifoldResult)
remove the mConstraintRow[3] from btManifoldPoint, it just took a lot of memory with no benefits: fixed it in btParallelConstraintSolver
For OSX:
cd build
./premake_osx xcode4
for iOS:
cd build
./ios_build.sh
./ios_run.sh
Also integrated the branches/StackAllocation to make it easier to multi-thread collision detection in the near future. It avoids changing the btCollisionObject while performing collision detection.
As this is a large patch, some stuff might be temporarily broken, I'll keep an eye out on issues.
Serialization: remove obsolete autogenerated headers
Minor changes in btSequentialImpulseConstraintSolver: split methods to make it easier to derive from the class and add functionality.
Tweak the BenchmarkDemo a bit:
1) disable deactivation in graphical mode
2) add some settings that increase performance in the BenchmarkDemo2 (1000 stack) from 35ms to 15ms on this quad core (at the cost of a bit of quality)
Also comment-out some code for __SPU__ to reduce code size
Added btContactConstraint (only used on PS3 SPU right now, better to use btPersistentManifold directly for contact constraints)
Improved readblend utility library (see also usage in http://gamekit.googlecode.com with Irrlicht)
Fix for btConvexConvexAlgorithm, potential division by zero
Thanks linzner http://code.google.com/p/bullet/issues/detail?id=260
Added optional flag btSoftBody::appendAnchor( int node,btRigidBody* body, bool disableCollisionBetweenLinkedBodies=false), to disable collision between soft body and rigid body, when pinned
Added btCollisionObject::setAnisotropicFriction, to scale friction in x,y,z direction.
Added btCollisionObject::setContactProcessingThreshold(float threshold), to avoid collision resolution of contact above a certain distance.
Avoid collisions between static objects (causes the CharacterDemo to assert, when a dynamic object hits character)
1) Add fast branchless SIMD support for constraint solver (Windows only until we get other contributions).
See resolveSingleConstraintRowGenericSIMD in Bullet/src/BulletDynamics/ConstraintSolver/btSequentialImpulseConstraintSolver.cpp
resolveSingleConstraintRowGenericSIMD can be used for all constraints, including contact, point 2 point, hinge, generic etc.
2) During this refactoring, all constraints support the obsolete 'solveConstraintObsolete' while we add 'getInfo1' and 'getInfo2' support.
This interface is almost identical interface to Open Dynamics Engine, to make it easier to port Dantzig LCP solver.
3) Some minor refactoring to reduce huge constructor overhead in math classes.
+ improved friction warm starting
+ made constraint solver configuration more consistent (moved m_solverMode into btContactSolverInfo)
+ reset timing in CDTestFramework after initialization (SAP init destorts timings)
+ make it easier to change default sizes for stack allocator in btDefaultCollisionConfiguration
- added bt32BitAxisSweep3, which co-exists without recompilation, using template class. This broadphase is recommended for large worlds with many objects (> 16384), until btMultiSwap is finished.
- Fixed some recent issues in Bullet 2.57 related to compound (thanks Proctoid) and memory allocations
