Bart/bullet3
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
2,810 Commits 10 Branches 0 Tags
c782f4976c73021a9ec13a9f3977a34c07c23ab1
Commit Graph

106 Commits

Author SHA1 Message Date
Jackson Lee
c782f4976c Various minor PLBVH related changes. 
-Use most significant bit instead of negative for internal nodes.
-Explicitly store root node index, so that it does not have to be 0.
-Check the root node first in PLBVH traversal.
-Fix rigid body clipping in RaytracedShadowDemo.
 2014-02-24 23:50:20 -08:00
Jackson Lee
e4fbd5332d Accelerate GPU raycaster with PLBVH. 2014-02-23 20:40:58 -08:00
Jackson Lee
b7b7356af8 Draft Parallel Linear BVH Broadphase. 2014-02-18 19:23:25 -08:00
erwincoumans
122ceacb6d move OpenCL initialization for the unit tests in a shared header file, and support some basic command-line arguments 
--cl_device=1 --cl_platform=1 --allow_opencl_cpu
add chaindemo, test for mass ratios
restore sleeping/activation mode in featherstone demo
Use _VARIADIC_MAX=10 to avoid Google Test issues with Visual Studio 2012, thanks to Mobeen for the report
Enable verbose printf for unit tests
 2014-02-11 10:33:00 -08:00
erwincoumans
bd5c2ff5ec batching reports the size of the batches, and solver uses this as termination condition, giving a good speedup 2014-02-07 06:28:15 -08:00
erwincoumans
3e8b183587 Avoid breaking up the clipHullHull kernel, it ruins performance. Unfortunately, Mac OSX still requires it. 
Use indices instead of copies for small/large aabbs in broadphase (grid / sap)
 2014-01-29 15:20:20 -08:00
erwincoumans
ff051f87aa replace 'copy' step of large/small Aabbs by indices 2014-01-29 12:52:31 -08:00
erwin coumans
33ebebd1c9 allow PairBenchmark to select broadphase type, with cpu brute-force, gpu brute-force etc 
fix issue in PairBench, related to index offset
allow to add a large AABB in PairBench, to see the effect on the broadphase pair search performance
 2014-01-28 17:11:56 -08:00
Erwin Coumans
71f0537c6e add Bullet 2.x constraint solver tests for hinge, with large mass ratio 
experiment with Midi controls in PairBenchmark to tune variables (will move it into a utility class, so every demo can easier use it in the future)
fix issue with broadphase, not initializing all pairs properly at pair array overflow
 2014-01-28 10:25:04 -08:00
Erwin Coumans
e3ee9e5b2e fix Mac OSX build, due to conflict empty #define __global and STL 2014-01-26 12:08:04 -08:00
erwincoumans
7cdda65822 Experimenting with GUI in PairBench, still preliminary. Goal is to make it easier to add variables that can be tuned using GUI/gwen 
Pre-compile PairBench OpenCL kernel
 2014-01-26 11:27:20 -08:00
erwin coumans
ec4292ad75 add basic test texture 2014-01-17 22:19:12 -08:00
Erwin Coumans
edb0c72166 enable MPR OpenCL for Mac OSX 2014-01-16 14:24:18 -08:00
Erwin Coumans
b53cc5edb8 disable Mpr on Mac OSX (need to split the kernel again...) 2014-01-15 11:24:35 -08:00
Erwin Coumans
472a4068fc Enable MPR by default 
Add the contact point from MPR, in addition to SAT/clipping contacts. Added a new kernel to clear/reset the number of contacts in pairs (stored in the z component)
Always sample unit sphere directions, if there are more edge-edge combinations than unit sphere directions (162 by default)
Remember last running demo for Bullet 3 (and save it in a text file, Bullet
Enable the testFileFracture.bullet in the Bullet2FileDemo
 2014-01-15 10:35:51 -08:00
erwin coumans
2e55d03fd9 more fixes towards working convex-convex, in case of (too) many edge-edge cases: 
1) sample directions over a unit sphere and
2) add a contact, using mpr
add missing bullet.pc.cmake file (for old Bullet 2.x)
 2014-01-09 16:22:34 -08:00
erwincoumans
271f458837 Ported Minkowski Portal Refinement mpr.c from libccd to OpenCL, for bettwe edge-edge performance (and additional contact point for degenerate/high detailed convex shapes) 
Removed b3RigidBodyCL, replace by b3RigidBodyData and b3RigidBodyData_t shared between C++ host and OpenCL,
Same for b3InertiaCL -> b3InertiaData
 2014-01-04 20:54:27 -08:00
erwin coumans
12dd0daebc prepare for Bullet 2 demo collection 2013-12-19 22:14:37 -08:00
erwin coumans
9ebef9552b implement 'new contact reduction' kernel on CPU , for debugging 2013-12-14 10:54:25 -08:00
erwin coumans
3fe969c4ee b3Solver -> pass pointer to source instead of 0 (was left over from a debugging session), thanks to David for the report 
Break up clipHullHullConcaveConvexKernel into multiple stages, so it might 'fit' in Apple's OpenCL implementation
Implemented bvhTraversalKernel and findConcaveSeparatingAxis on CPU (debugging, possible future CPU version)
 2013-12-13 07:52:41 -08:00
erwin coumans
c155e126d0 move parts of collision pipeline to shared header files (work-in-progress) 2013-12-12 11:03:55 -08:00
erwin coumans
beb9e98cd3 expose the gUseLargeBatches as command-line argument --use_large_batches 2013-11-21 09:15:52 -08:00
erwincoumans
1d5c651753 Use more shared data structures in src\Bullet3OpenCL\NarrowphaseCollision\kernels\satClipHullContacts.cl 
Revert default error func (was broken)
 2013-11-19 17:52:59 -08:00
erwin coumans
26dfaa441e add some command-line options 
--use_jacobi
--allow_opencl_cpu

b3LauncherCL constructor takes string, to make it easier to determine failing OpenCL kernel
b3SetCustomErrorMessageFunc, printf error and exit(0)
 2013-11-19 13:42:53 -08:00
erwincoumans
472614772f expose some more settings as command-line parameters: 
[--disable_shadowmap] [--shadowmap_size=int] [--shadowmap_resolution=<int>] [--use_jacobi] [--use_uniform_grid]
 2013-11-11 22:20:21 -08:00
erwincoumans
5ce0b3938f Re-introduce GPU parallel Jacobi solver, called btGpuJacobiContactSolver. There are still some issues, but for basic scene's it works. 
To avoid confusion, rename GPU contact solver to btGpuContactSolver, and constraint (non-contact) solver to btPgsConstraintSolver.
 2013-11-11 21:15:06 -08:00
erwincoumans
ef6be5370d more work towards hybrid of GPU grid and sap broadphase, separating small, large (moving) and static objects 
re-enable 'useNewBatchingKernel', it is slower but more robust
 2013-11-11 02:19:04 -08:00
erwin coumans
d13de109a5 work towards re-enabling GPU grid broadphase (in addition to GPU sap bp) 2013-11-10 23:19:51 -08:00
erwincoumans
b0c43520e3 more work towards re-enabling grid broadphase. 2013-11-10 22:00:35 -08:00
erwincoumans
ad8585f184 prepare for btGpuGridBroadphase 2013-11-10 14:15:23 -08:00
erwincoumans
8f3051fdf1 add ImplicitCloth demo, by Stan Melax 
add SimpleOpenGL3App, used in ImplicitCloth demo. The Bullet 3.x demos should use this as a template, it will clean up some of the mess.
 2013-11-09 12:12:33 -08:00
erwincoumans
a6bc61ad1a tweaked the particle demo a little bit 2013-11-08 23:21:01 -08:00
erwincoumans
768ea211a6 add a waiting message, during startup of a demo. 
fix an OpenCL kernel, broken in previous commit.
 2013-11-07 16:47:18 -08:00
erwin coumans
e85bae5fe3 share more data structures and code between OpenCL kernels and C/C++ code on CPU (non-OpenCL) 
integrateSingleTransform
fix bug in registerRigidBody (could lead to random crashes, especially when performing picking/adding rigid bodies afterwards)
 2013-11-07 15:49:57 -08:00
erwin coumans
9d96c24bee fix a compile error in the debug/testing particles kernel 2013-11-04 23:46:03 -08:00
erwincoumans
8faac34801 move ExplititEuler cloth solver to its own file, so it can be replaced by PDB, implicit euler etc 
add wireframe toggle ('w' key)
add --paused command-line option to start simulation in paused mode
render cloth using drawPoints method
 2013-09-06 18:09:35 -07:00
erwincoumans
181a323631 move b3RaycastInfo.h so it can be used without OpenCL dependency 
add b3Assert to make sure m_maxShapeCapacityInBytes is not exceeded in GLInstancingRenderer
prepare for CpuSoftClothDemo
 2013-09-02 22:31:48 -07:00
erwincoumans
0ccb6922a8 more work towards CPU version 2013-08-30 08:06:46 -07:00
erwincoumans
ad8d0e32f4 expose 2 new command-line options [--disable_cached_cl_kernels] [--load_cl_kernels_from_disk] 
on Windows (_WIN32) debug output goes to BOTH console and Visual Studio output log window.
 2013-08-26 21:58:55 -07:00
erwincoumans
d32629dbe2 add the option to load a set of AABB's from a file, see PairBench 
You can use the following command-line arguments:
App_Bullet*.exe --selected_demo=10 --cl_platform=1 --pair_benchmark_file=64006GPUAABBs.txt --benchmark
 2013-08-23 17:55:28 -07:00
erwin coumans
f12fe0237b Merge branch 'master' of https://github.com/erwincoumans/bullet3 2013-08-22 23:16:02 -07:00
erwin coumans
2c019d579f reduce memory allocation, lack of GPU memory totally destroys performance on my GTX 650M on the Macbook retina 
add some keys to toggle gui drawing and shadows
 2013-08-22 23:15:37 -07:00
Erwin Coumans
74948ede54 fix Linux build 2013-08-22 17:49:59 -07:00
Erwin Coumans
d860e7f51a fix Mac OSX build (compound-compound still broken on OSX) 
remove operator= from b3Vector3 to make it POD/Plane Old Data
 2013-08-20 12:45:05 -07:00
erwincoumans
677722bba3 support compound versus compound collision shape acceleration on GPU, using aabb tree versus aabb tree. 
Remove constructor from b3Vector3,  to make it a POD type, so it can go into a union (and more compatible with OpenCL float4)
Use b3MakeVector3 instead of constructor
Share some code between C++ and GPU in a shared file: see b3TransformAabb2 in src/Bullet3Collision/BroadPhaseCollision/shared/b3Aabb.h
Improve PairBench a bit, show timings and #overlapping pairs.
Increase shadowmap default size to 8192x8192 (hope the GPU supports it)
 2013-08-20 03:19:59 -07:00
erwincoumans
906415429c added (and stripped) a simple C/C++ preprocessor (written in Lua), so the stringifier can handle the #include directive, 
and embed the included files directly in the stringified files.
We need this, because we start sharing struct definitions and code between C/C++ and OpenCL (and potentially other languages)
preprocessor is from http://github.com/willsteel/lcpp
 2013-08-01 21:05:19 -07:00
erwincoumans
34de49d8a4 add gjk/epa (host only), possibly improve convex-convex with many edge-edge tests 
more preparation towards persistent/incremental contact cache
 2013-07-31 23:22:43 -07:00
erwincoumans
7992ff816b use int4 for broadphase pair, it allows to store persistent information in the future 
(contact cache, applied impulse/warm starting info etc)
 2013-07-31 09:58:15 -07:00
erwincoumans
338118b3c6 expose the number of frame that the benchmark mode runs as commandline parameter: 
for example --max_framecount=1000
 2013-07-30 18:00:14 -07:00
erwincoumans
1c0fc6bc2d report when compound pair capacity is exceeded (and avoid crash) 
add tetrahedron, reorganized demos
 2013-07-30 17:00:50 -07:00