CCD dataclass for ccd memory

[thowell]

the current testspeed memory report does not include memory allocated for convex collision detection (ccd)

in order to provide memory information for ccd this pr:

• adds dataclass CCD containing memory for ccd
• updates testspeed memory report: change utilized memory to be the sum of Model and Data memory since wp.get_mempool_used_mem_current and wp.get_mempool_used_mem_high do not include ccd memory allocated in collision_convex.convex_narrowphase (i think the memory is released after the function call?)

this update retains the property that memory is not allocated for ccd if collision_convex.convex_narrowphase is never called. however, if the function is called the ccd memory is no longer released after the call but is persistent in Data.ccd.

---

aloha + pot

mjwarp-testspeed ./benchmark/aloha_pot/scene.xml --nconmax=24 --njmax=128 --memory=True

this pr:

Rolling out 1000 steps at dt = 0.002...

Summary for 8192 parallel rollouts

Total JIT time: 0.68 s
Total simulation time: 4.95 s
Total steps per second: 1,656,290
Total realtime factor: 3,312.58 x
Total time per step: 603.76 ns
Total converged worlds: 8192 / 8192

Total memory: 2002.56 MB / 48640.12 MB (4.12%)
Model memory (0.27%):
 (no field >= 1% of utilized memory)
Data memory (99.73%):
 geom_xmat: 57.38 MB (2.87%)
 efc.J: 96.00 MB (4.79%)
 ccd.epa_vert: 90.00 MB (4.49%)
 ccd.epa_vert1: 90.00 MB (4.49%)
 ccd.epa_vert2: 90.00 MB (4.49%)
 ccd.epa_vert_index1: 30.00 MB (1.50%)
 ccd.epa_vert_index2: 30.00 MB (1.50%)
 ccd.epa_face: 407.25 MB (20.34%)
 ccd.epa_pr: 407.25 MB (20.34%)
 ccd.epa_norm2: 135.75 MB (6.78%)
 ccd.epa_index: 135.75 MB (6.78%)
 ccd.epa_map: 135.75 MB (6.78%)

main (528695a):

Rolling out 1000 steps at dt = 0.002...

Summary for 8192 parallel rollouts

Total JIT time: 0.69 s
Total simulation time: 4.95 s
Total steps per second: 1,654,766
Total realtime factor: 3,309.53 x
Total time per step: 604.31 ns
Total converged worlds: 8192 / 8192

Total memory: 432.81 MB / 48640.12 MB (0.89%)
Model memory (1.24%):
 (no field >= 1% of utilized memory)
Data memory (98.76%):
 xfrc_applied: 4.88 MB (1.13%)
 xmat: 7.31 MB (1.69%)
 ximat: 7.31 MB (1.69%)
 geom_xpos: 19.12 MB (4.42%)
 geom_xmat: 57.38 MB (13.26%)
 site_xmat: 4.78 MB (1.10%)
 cinert: 8.12 MB (1.88%)
 actuator_moment: 10.06 MB (2.32%)
 crb: 8.12 MB (1.88%)
 qM: 18.00 MB (4.16%)
 qLD: 16.53 MB (3.82%)
 cvel: 4.88 MB (1.13%)
 cacc: 4.88 MB (1.13%)
 cfrc_int: 4.88 MB (1.13%)
 cfrc_ext: 4.88 MB (1.13%)
 contact.frame: 6.75 MB (1.56%)
 contact.efc_address: 4.50 MB (1.04%)
 efc.J: 96.00 MB (22.18%)
 efc.quad: 12.00 MB (2.77%)
 subtree_bodyvel: 4.88 MB (1.13%)

[hartikainen]

Use default values or __post_init__?

[hartikainen]

isinstance accepts tuple as the second argument.

Suggested change

	if any(isinstance(value, mjw_type) for mjw_type in (mjw.Contact, mjw.Constraint, CCD)):
	if isinstance(value, (mjw.Contact, mjw.Constraint, CCD)):

[thowell]

thank you for the review @hartikainen!

[thowell]

closing in favor of arena memory #966