Sam Haves
799f8a300e
Add parsing benchmark simulating many textures.
...
PiperOrigin-RevId: 857883857
Change-Id: Ia428ff8aeba4fbaae67a6337d0856cd2b19af2e6
2026-01-18 11:17:45 -08:00
Yuval Tassa
849a650143
Refactor mju_cholUpdateSparse to use a dense accumulator, 1.5-2x function speedup.
...
PiperOrigin-RevId: 846754379
Change-Id: Ie599658db2907a26095072d97d07b99a693050dc
2025-12-19 08:57:45 -08:00
Yuval Tassa
45b0153067
Refactor sparse Cholesky factorization into symbolic and numeric phases.
...
The new symbolic function is a generalization of the function it replaces. In this CL it takes two unused temp arrays. The actual change in behavior happens in the followup.
New benchmark test output below ("L" is 2 humanoids and 100 free objects, "XL" is 100 humanoids). Note that `symbolic` is only ever called once per Newton iteration, while `numeric` is sometimes called multiple times (when the rank-1 update fails), hence timing them separately is valuable.
```
Benchmark Time(ns) CPU(ns) Iterations
--------------------------------------------------------------
BM_old_L_mean 84382 84703 19547 11.807k items/s
BM_symbolic_L_mean 16345 16381 88414 61.055k items/s
BM_numeric_L_mean 10986 10994 120000 90.999k items/s
BM_old_XL_mean 1241208 1244212 1200 803.924 items/s
BM_symbolic_XL_mean 130917 131042 12720 7.631k items/s
BM_numeric_XL_mean 77004 76767 21116 13.029k items/s
```
PiperOrigin-RevId: 846704054
Change-Id: Ib0c365724d63bf2b81606ca5353756a6496c3a26
2025-12-19 06:11:55 -08:00
Yuval Tassa
769f37b653
Implement sleeping in engine
...
PiperOrigin-RevId: 829361787
Change-Id: I6f64d8e25c4248cf32c18cd94d37ff5def78946e
2025-11-07 03:33:07 -08:00
Yuval Tassa
ab9575d8a6
Refactor and speed up mj_crb and add a benchmark.
...
This is a no-op "prefactor" of `mj_crb` to reduce the number of lines changed in the upcoming sleeping CL. The main change here is that `mj_crb` now avoids accessing the model and data pointer repeatedly, but instead has the input and output pointers explicitly declared as local variables. A benchmark test found a healthy **14% perf bump** due to two changes:
- Adding `restrict` to `mju_mulInertVec`
- The function-local pointers.
Adding `restrict` to the local pointers had no effect. Note that `mj_crb` is not a particularly expensive function so these speed bumps are not significant per se, but rather indicative of possible future gains with these techniques.
```
Benchmark Time(ns) CPU(ns) Iterations
---------------------------------------------------------------
ORGINAL BASELINE
BM_CRB_BASELINE_mean 4325 4348 993200 230.052k items/s
BASELINE + RESTRICT
BM_CRB_BASELINE_mean 4057 4090 1186150 244.573k items/s
LOCAL POINTERS + RESTRICT
BM_CRB_mean 3772 3800 1001750 263.209k items/s
```
PiperOrigin-RevId: 815798225
Change-Id: Iffdf57b544e8e10562807c617f73ca4ccd1c614f
2025-10-06 10:54:01 -07:00
Kyle Bayes
91257896c0
Implement faster mju_isZeroByte.
...
PiperOrigin-RevId: 814148714
Change-Id: Iaf5ed9d3502eaaaee8d46a986f2db00811ae0707
2025-10-02 02:53:53 -07:00
Sam Haves
c31f11824e
Add hfields to cache_hits benchmark.
...
PiperOrigin-RevId: 812759213
Change-Id: I155c2db58811dd1efa56a6361fd39aac669b63e7
2025-09-29 06:22:24 -07:00
Sam Haves
e3377ceca3
Add cache hit benchmark to parsing benchmarks.
...
PiperOrigin-RevId: 810495543
Change-Id: I8c10806a091f870266f2c1194b39159c1f3105e0
2025-09-23 10:40:12 -07:00
Yuval Tassa
d71d42a374
Add private function mju_isZeroByte to check for byte-wise zero.
...
PiperOrigin-RevId: 810456282
Change-Id: I877d0f9225e7a783ed7ff12ce54f4ed0a7889793
2025-09-23 08:49:43 -07:00
Yuval Tassa
681f5767b2
Remove -Wno-extra-semi, add workaround for mjSORT confusing some IDEs
...
PiperOrigin-RevId: 808689636
Change-Id: Ia091fb597c171a33842e62690d06f32614fbe4eb
2025-09-18 12:28:09 -07:00
Yuval Tassa
ec94bb49aa
Make constraint island discovery on by default.
...
Also fix latent bug in mjData serialization.
PiperOrigin-RevId: 799505349
Change-Id: I299e7cc8133fa2ec0b339a7ff3d02543e06778b1
2025-08-26 04:27:24 -07:00
Yuval Tassa
2f28473bc1
Remove transposed Jacobian fields from mjData
...
No longer computed unconditionally, only ad-hoc where required.
PiperOrigin-RevId: 797753297
Change-Id: I2030fc342c98ff33575b0526dd72e2110c4fcb74
2025-08-21 07:21:24 -07:00
Yuval Tassa
977f94e9df
Delete sparsity structures from mjData, use the ones in mjModel
...
PiperOrigin-RevId: 795067950
Change-Id: If4f52e40ac49f7c6986b33ef9c8b534199dfd21a
2025-08-14 09:13:02 -07:00
Yuval Tassa
365cac49f8
Add benchmarks for mjSORT macro.
...
PiperOrigin-RevId: 781555531
Change-Id: Ia33b5512fbf65b08931037859d8c6062b018d760
2025-07-10 09:00:46 -07:00
Alessio Quaglino
7edbdd0ad6
Remove the Shell plugin and integrate it into the engine.
...
PiperOrigin-RevId: 760688502
Change-Id: Ia70988d42b7edf571d7cb4a4f48f8fc50b51667d
2025-05-19 10:51:25 -07:00
Yuval Tassa
45fc15b844
Speed up sparse supernode detection by combining it with transposition.
...
PiperOrigin-RevId: 760673008
Change-Id: I6d66580e675fd86e5b974859383f93f87482ca16
2025-05-19 10:14:15 -07:00
Yuval Tassa
ee8abdf854
Speed up mju_sqrMatTDSparse
...
PiperOrigin-RevId: 759593417
Change-Id: I0168e1f96333769d09d61e835560910d43aac608
2025-05-16 06:42:17 -07:00
Yuval Tassa
d8bebdc675
Remove unnecessary diagnum argument in mj_solveLD and mj_factorI
...
PiperOrigin-RevId: 758669364
Change-Id: Icfefc4a7a1e7d28da373725358ccf9d5f589f689
2025-05-14 07:13:27 -07:00
Yuval Tassa
dd28b887d4
Rename C sparse structure to M in mjdata, improve docstrings.
...
PiperOrigin-RevId: 758636638
Change-Id: If78acc423601d2911f514929b27f7b6d0af9ef58
2025-05-14 05:26:19 -07:00
Yuval Tassa
627fffdef9
Switch mjData.{qH,qLD} from full ("M") to reduced ("C") inertia matrix structure.
...
PiperOrigin-RevId: 758273074
Change-Id: If1a2e663ea70044694af985e0119afd6d58115ac
2025-05-13 10:20:32 -07:00
Yuval Tassa
b53ed63b5f
Use 2humanoid100.xml model in a benchmark test.
...
PiperOrigin-RevId: 757784271
Change-Id: I7da11ab85ba9f5e023cdbe3c77d31224f3444ca5
2025-05-12 08:59:14 -07:00
Yuval Tassa
c2ac0d724e
Add mju_gather and mju_scatter, private engine functions.
...
PiperOrigin-RevId: 750245294
Change-Id: I3a14ccdd55a324d3fe206e388fb2a513772f5f5e
2025-04-22 10:33:33 -07:00
Kyle Bayes
606f00f802
Remove contact pruning with box-box collisions in nativeccd and add benchmarks.
...
PiperOrigin-RevId: 744732382
Change-Id: I2e20b646c541ee99888f1e964302d1b4363b2dc4
2025-04-07 08:34:04 -07:00
Kyle Bayes
467a63d203
Add BoxBox to ccd benchmarks.
...
PiperOrigin-RevId: 738761460
Change-Id: If59d306748d64e9e30abe65daa746a4f8e9ec78c
2025-03-20 05:16:33 -07:00
Yuval Tassa
0f563ecf31
Switch mjData.{qH,qLD} from reduced ("C") to full ("M") inertia matrix structure. No performance impact of extra zeros because of existing "simple dof" skipping mechanism.
...
PiperOrigin-RevId: 733523931
Change-Id: Ic8d8a152dda5532331c239cb6b4ce7d8d09b7fff
2025-03-04 17:34:35 -08:00
Kyle Bayes
ed16f2daf2
Set nativeccd as default.
...
PiperOrigin-RevId: 731343200
Change-Id: I315779017b676f3d118e0d38ee53a515272f50b8
2025-02-26 09:15:15 -08:00
Alessio Quaglino
c52d1b3941
Remove the particle composite (replaced by replicate).
...
PiperOrigin-RevId: 730089997
Change-Id: If8d18fb56ae01241e76f84dbf7951e034026d829
2025-02-23 03:14:11 -08:00
Yuval Tassa
b516edae1b
Rename inertia factorization routines:
...
```
mj_factorI -> mj_factorI_legacy
mj_solveLD -> mj_solveLD_legacy
mj_factorIs -> mj_factorI
mj_solveLDs -> mj_solveLD
```
PiperOrigin-RevId: 728246367
Change-Id: I3bc7804cb96faac2ae5419ba9abfc3c1648ed4c6
2025-02-18 09:42:05 -08:00
Yuval Tassa
fb07e0d1f7
Convert qLD to CSR format.
...
PiperOrigin-RevId: 728196818
Change-Id: I31bdf32e251293284426ed345019f27fd051a0e1
2025-02-18 07:14:45 -08:00
Yuval Tassa
e39a5df06a
mju_compressSparse(): add option to compress away small elements.
...
PiperOrigin-RevId: 726010576
Change-Id: Ib6458037a5cf9a9d82abb364b7dda7aab2f36807
2025-02-12 05:08:20 -08:00
Yuval Tassa
5ee0654830
Roll back recent change to mjData.qLD until some issues are resolved.
...
PiperOrigin-RevId: 724389161
Change-Id: I1b30ab0950ab5c5e611b1c6fc3bf3f42f964939c
2025-02-07 10:21:42 -08:00
Yuval Tassa
c27d3758c2
Convert qLD to CSR format.
...
PiperOrigin-RevId: 723955038
Change-Id: I30c3dc7f59739e89ae5fff8841432bc74717ec1b
2025-02-06 08:59:35 -08:00
Yuval Tassa
00010f5848
Compute diagonal indices in mj_sqrMatTDSparse
...
PiperOrigin-RevId: 713714087
Change-Id: Icc8eae74e6e47ba1d12a6e9aa0d774ab006cfe38
2025-01-09 10:17:31 -08:00
Yuval Tassa
8a5f092081
Allow CSR back-substitution to handle multiple vectors.
...
PiperOrigin-RevId: 713229673
Change-Id: I7a5b43fe966cf9e482bd41e2eea6c30dd3ffa1d4
2025-01-08 03:32:16 -08:00
Yuval Tassa
ac11e5faa6
Add CSR implementation of mj_factorI
...
PiperOrigin-RevId: 712498431
Change-Id: I13b52e53482ed97da8788875d4d95e2beb5ca7c1
2025-01-06 05:48:14 -08:00
Yuval Tassa
7eb8231fda
Represent only the lower triangle in Newton solver's reduced dof-dof matrix.
...
PiperOrigin-RevId: 712488529
Change-Id: Iad91c72654376539791d7856765a0d0ac9088251
2025-01-06 04:58:55 -08:00
Yuval Tassa
4510c6d290
Further speedup to CSR back-substitution using dof_simplenum.
...
PiperOrigin-RevId: 711440268
Change-Id: I81cd9a6a8b8ec78d08cfbc34f60a9216ea753833
2025-01-02 09:04:51 -08:00
Yuval Tassa
ea98c57921
Clean up engine_core_smooth_benchmark_test.
...
PiperOrigin-RevId: 709573203
Change-Id: I461f259c33d74b3eaf0b6fe58124fe09418c5303
2024-12-25 07:44:25 -08:00
Copybara-Service
526637f118
Merge pull request #2276 from traversaro:reducetestboilerplate
...
PiperOrigin-RevId: 707613522
Change-Id: Ic945c6b057b77f3462dbdb758139af12133f6a8f
2024-12-18 11:22:59 -08:00
Yuval Tassa
1c4c7b012c
CSR implementation of mj_solveLD.
...
PiperOrigin-RevId: 704316200
Change-Id: Ibaff0284e40b3ebbe43bb489b6211ce739270e27
2024-12-09 09:34:17 -08:00
Silvio
756a8d716a
Reduce cmake test boilerplate
...
All instances of mujoco_test in the code base were followed by
a call to target_link_libraries(<testname> fixture gmock).
As anyhow mujoco_test already was calling target_link_libraries
to some predefined list of targets (mujoco and gtest_main),
this PR adds to the list of default linked targets also fixture
and gmock, to reduce the boilerplate.
Furthermore, to completly remove the need for calling
target_link_libraries after a call to mujoco_test, this PR also
add to the mujoco_test macro the ADDITIONAL_LINK_LIBRARIES
argument, that can be used if a given test needs to link some
additional targets beside the default ones.
To permit to use these new features also for mujoco benchmarks,
this PR adds a MAIN_TARGET parameter to mujoco_test, to select
if gtest_main or another target is used to provide the main
entry point to the test executable.
2024-12-07 17:14:20 +01:00
Yuval Tassa
a4a6248a06
Move AsVector utility to fixture.h
...
PiperOrigin-RevId: 690577546
Change-Id: I2cebcffa1f2e3b0789f43e772e364c4762719ab3
2024-10-28 05:52:25 -07:00
Yuval Tassa
62ec802ec1
Tweaks to benchmarks.
...
- Modify parse_benchmark_test to not time the first instance of file loading.
- Improve readability of step_benchmark_test.
PiperOrigin-RevId: 680953562
Change-Id: I4569d1b946ecb49b6590ea313f0b598ffc897888
2024-10-01 04:08:40 -07:00
Kyle Bayes
43a82e08be
Add benchmark test for collisions.
...
PiperOrigin-RevId: 680554813
Change-Id: I0be5b102d68aa1a54b4ceca6b2bd24fbc519f5df
2024-09-30 06:46:25 -07:00
Yuval Tassa
4e3dc14f8e
Remove model/humanoid100/humanoid100.xml (used <include>), replace with model/humanoid/humanoid100.xml (uses <attach>)
...
PiperOrigin-RevId: 673123895
Change-Id: I0c3c6dac69d9fc12a6ea76d586c34503fc167bad
2024-09-10 15:14:04 -07:00
Alessio Quaglino
4862b9e765
Remove plugin support from Composite particles.
...
This functionality is replaced by FlexComp.
PiperOrigin-RevId: 672503965
Change-Id: I58953267de285f82e416f9f60576a336b7d0da4c
2024-09-09 04:58:26 -07:00
Yuval Tassa
19d505d847
Remove unused texture attribute in some models.
...
PiperOrigin-RevId: 648497401
Change-Id: I9f9a843abfb710afaf25798fa659efd92bcd7007
2024-07-01 15:05:02 -07:00
Kyle Bayes
b47d97652e
Load from VFS in parsing benchmark.
...
PiperOrigin-RevId: 610384849
Change-Id: Id226c30f1629d6b4bbc2231c1e998e733c27772e
2024-02-26 06:08:11 -08:00
Alessio Quaglino
83515446b6
Add missing models to all_models and all_test_models.
...
PiperOrigin-RevId: 574491634
Change-Id: I943043b21056d7f22dd1c50454deee3f49e7eac0
2023-10-18 08:59:54 -07:00
Alessio Quaglino
6f8128dc6a
Fix benchmark label for plugin flag test.
...
PiperOrigin-RevId: 572891541
Change-Id: I442e31f7e4164789759be0d2c71275fdb1398955
2023-10-12 06:53:38 -07:00