Every step, the flex block of the implicit effective metric M + K was
factorized by sparse Cholesky, because K depends on the configuration. On
model/flex/bag.xml, added here, that is roughly half the step, against a
comparable share for the constraint solve it exists to accelerate.
Keep only the metric's per-vertex 3x3 diagonal blocks, prefactored. Neither
consumer needs the exact inverse: the CG constraint solver only wants a
preconditioner, and qacc_smooth can come from an iterative solve using those
blocks. They are O(n) to build and to apply, but weaker, so CG runs about twice
the iterations and qacc_smooth becomes an iteration rather than a direct solve.
Net, the bag model steps roughly twice as fast.
The preconditioner, by metric state. Inactive, meaning no flex elasticity or an
explicit integrator: M^-1, unchanged. Bending only (nefmK == 0): M^-1 plus the
exact constant bending factor from mj_setConst on the dofs it covers,
unchanged; that factor is built at model compile time and costs nothing per
step. Per-step stiffness: M^-1 plus the 3x3 blocks, where before it was a
per-step sparse Cholesky, or, when M couples across the flex block, an inner
PCG of up to 50 iterations run once per outer CG iteration.
Only models carrying per-step stretch stiffness change in wall-clock. Both
ponchos hold their timing and take slightly fewer CG iterations than before,
because the preconditioner is now symmetric: it applies M^-1 and the covered
blocks to disjoint sets of dofs, where previously the two overlapped and the
operator was not symmetric, which PCG requires.
mjd_effSolve is the accurate solve of (M + K)x = b; what used to carry that
name only preconditions and is now mjd_effPrec. Its CG guarded the division by
pAp with mjMINVAL, an absolute floor on a quantity that scales with the square
of the right-hand side, so a small b aborted the solve while the curvature was
healthy: four flex models were quietly left short of tolerance. For an SPD
metric the guard is positivity, and with that the same solves converge. The qacc_smooth call site in
mj_fwdAcceleration is textually unchanged but now reaches the iterative solve,
which converges on opt.tolerance rather than a hardcoded threshold, floored in
mjUSESINGLE builds where the squared target is unreachable in float. Reaching
the iteration cap names the ill-conditioned flex stiffness and then reports it
through mjWARN_INERTIA, rather than returning an under-converged result.
Covered dofs are located by walking the covered rows of the stiffness matrix,
as they need not be 3-aligned from dof 0: any joint declared before a flexcomp
shifts them.
mjData.efm_L_rownnz, efm_L_rowadr and efm_L_colind described the sparsity of
the deleted factorization and are removed: left NULL with nonzero mjxmacro
extents they made the Python bindings hand back uninitialized arrays.
efm_active loses the value 2 for the same reason, nothing selects a solve path
on preconditioner exactness any more. Both are recorded under breaking changes.
model/flex/bag.xml is added because no shipped model carried per-step stretch
stiffness. The ponchos are bending-only and trampoline.xml uses an explicit
integrator, so the metric never activates there. It is excluded from
WriteReadCompareTest: stretch stiffness amplifies rest geometry that XML rounds
on save.
This CL replaces the post-hoc implicit flex correction (`flexInterp_cgsolve`) with a **linearly-implicit effective metric** `M̃ = M + (h² + h·damping)·K` carried by the CG constraint solver itself. Contact/friction forces and implicit flex elasticity are now computed against one consistent metric, instead of the solver seeing `M` and a post-solve correction changing `qacc` behind its back.
Gate (unchanged semantics): `solver="CG"` + implicit/implicitfast integrator + pyramidal cones + flex stiffness present. Newton and PGS are untouched. `solver="CG"` remains the user-facing contract — the factorization is an implementation detail of the preconditioner.
### What's in the metric
- **mjData `efm_*`** (arena, efc-like lifetime/skip semantics; built in `mj_fwdPosition`, value-refreshed in `mj_fwdVelocity`): the per-step stiffness CSR `efm_B_*`, its reverse-Cholesky factor `efm_dofid` + `efm_L_*` (nested-dissection ordered, separators-first for the reverse factorization), and the smooth-force shift `efm_c = h·K·qvel`.
- **`mjd_flexStiff_assemble`** now assembles stretch (Gauss–Newton), standard dim-2 bending, and — via the cached corotated stiffness `d->flexelem_krot` — interp stiffness (all node bodies on simple sliders: point Jacobian is I₃, `flex_centered` not required; fixed nodes drop like pins) into one dof-level CSR. `mjd_effMulAdd`/`mjd_effSolve` apply the metric, with matrix-free operator fallbacks where assembly does not apply.
- **mjModel `efm0_*`** (`nefm0dof`/`nefm0L`): the constant part of the metric factor — currently the dim-2 bending factor, computed once in `mj_setConst` — so bending-only models pay zero per-step factorization cost. Naming mirrors mjData's `efm_*` with the standard `0`-suffix (reference/constant) idiom, and is deliberately not bending-specific: future constant contributors extend it without renames.
- The solver consumes the metric through pre-shifted `qfrc_smooth` and the metric products `Ma`/`Mv`/`Mgrad`; `qacc_smooth` becomes the unconstrained minimizer of the implicit dynamics, which makes the no-constraint shortcut and the warmstart choice consistent by construction.
- **`mj_inverse` adds `B·qacc − c`**, making inverse dynamics discrete-consistent with the gated forward dynamics — exact, since the gated path has no qDeriv term (new test `ForwardTest.GatedFlexInverseConsistency`).
### Performance
All numbers: ms/step over the same 2000-step window, models as shipped on each side (old code with the old model settings vs this CL with the new ones).
The new solver path activates on exactly two shipped models — the ponchos, the only flex models that need an implicit integrator (poncho on Euler degenerates to >200 ms/step). For them, this CL trades speed for consistency: the implicit bending solve now runs inside every solver iteration, where the contact solve can see the stiffness, instead of once after the solve. Solver iterations drop because the curvature is visible, but each iteration pays for the implicit solve:
| model | before | after | solver iters/step |
|---|---|---|---|
| poncho | 2.47 | 3.30 (1.33×) | 16.8 → 11.8 |
| poncho_edgeequality | 1.96 | 2.72 (1.39×) | 13.2 → 10.0 |
What that price buys: contact forces consistent with the implicit elasticity (previously the post-hoc correction changed `qacc` after the constraint solve), discrete-consistent inverse dynamics, and the removal of the post-hoc special case from the integration path. Raising poncho's timestep from 2 to 5 ms leaves its per-step cost nearly flat, so the consistency price can be recovered by taking fewer steps where accuracy allows.
Every other flex model was measured stable on Euler at its shipped timestep and switches to it (these models predate the post-hoc integrator; implicit was never load-bearing for them). They end up equal or faster than before: bunny_multicell 0.47 → 0.40, trampoline 0.28 → 0.25, plate 1.02 → 0.99, pancake 0.34 → 0.33.
Finally, the per-step factorization makes configurations practical that the old code could only integrate explicitly: implicit stretch elasticity (`elastic2d="stretch"`/`"both"`, dim-3 solids) and factorized interp stiffness. No before/after exists for these — stock has no implicit treatment of stretch at all.
### Behavior changes
- With the post-hoc correction deleted, interp/bending models running `solver="Newton"` (or elliptic cones, or islands) now integrate flex elasticity **explicitly** (previously: post-hoc implicit). Affects e.g. `gripper_trilinear` (stable, and faster, but different semantics). Follow-up options: Newton-side metric support, or a documented fallback.
- With the gate on, `mj_forward` outputs are timestep-dependent for gated models (they answer the linearly-implicit discrete problem); `qacc_smooth` and `mj_inverse` change accordingly. Non-gated models are bit-identical (full suite green throughout).
### Validation
- 1737/1737 tests, including new: `FlexStretchDerivatives` (FD-validated GN operator), `FlexStiffAssemble`/`FlexStiffAssembleInterp` (CSR ≡ operators), `GatedFlexInverseConsistency` (fails pre-change), equivalence tests vs the old post-hoc treatment (bending matches to 2e-11).
- Fingerprint discipline throughout: bending-only models bit-exact across every refactor; permutation/kernel changes verified iteration-identical.
### Known follow-ups (not in this CL)
3×3-block sparse Cholesky kernel (the numeric factorization is index-bound; projected ~3× on the factor); mjModel persistence of the factor's symbolic pattern (rest-pose ND makes sizes compile-time); the general effective-metric mode (all solvers, all PSD-safe force classes, behind an enable flag).
PiperOrigin-RevId: 948561856
Change-Id: I8b8e32ebd0428042af71647d0470d10773bf6daf
Extend mjpResourceProvider with an optional write callback (write)
so that mj_encode, mj_saveXML, and mj_saveModel can write to any
registered provider.
PiperOrigin-RevId: 945202741
Change-Id: I37903425260932e555f4a8c2392c4ff8c2e6cc06
The field now stores either an approximate or exact diagonal of the constraint matrix A, so the name is made more general.
PiperOrigin-RevId: 924244954
Change-Id: I62b2f76531fb88b7b3bf96e6769940197596702b
For example, `compile mujoco_menagerie/robotis_op3/scene.xml` now outputs
```
Compile 1 (cold cache):
total: 317.2 ms
assets: 284.8 ms (wall clock)
load: 616.1 ms
hull: 26.4 ms
poly: 137.8 ms
inert: 177.8 ms
bvh: 568.2 ms
octr: 1.6 ms
tex: 25.3 ms
other: 32.4 ms
Compile 2 (warm cache):
total: 79.9 ms
assets: 54.5 ms (wall clock)
load: 888.5 ms
hull: 0.0 ms
poly: 0.0 ms
inert: 0.0 ms
bvh: 0.0 ms
octr: 0.0 ms
tex: 21.4 ms
other: 25.3 ms
```
PiperOrigin-RevId: 917850214
Change-Id: Iaec86230bec0faf2e47820e20cbff61de5b2621e
Add a new mjpEncoder plugin type mirroring the existing mjpDecoder pattern.
Encoders serialize an mjSpec + mjModel to an mjResource for a given format.
New API functions:
- mjp_registerEncoder: globally register an encoder
- mjp_defaultEncoder: zero-initialize an encoder struct
- mjp_findEncoder: look up an encoder by filename extension or content type
The mjfEncode callback takes (mjSpec*, mjModel*, mjVFS*, mjResource*) and
returns 0 on success. Writing to mjResource keeps symmetry with the decoder
reading from mjResource and leaves the door open for writable resource providers.
PiperOrigin-RevId: 889187898
Change-Id: I180771b2255b91dea188ac5e2cdc3a8f0fb85364
This change adds an Emscripten binding for loading binary MuJoCo models (`.mjb`) from the virtual filesystem. The new binding, `mj_loadBinary`, takes a filename and an `MjVFS` object, enabling the loading of models that reference assets within the VFS.
PiperOrigin-RevId: 859740455
Change-Id: I9f9d8bbba0b57422e2356867e4282a1b04566475
Regarding mj_compile VFS argument:
I was encountering a memory error when using `std::optional<MjVFS>` with `mj_compile` (because vfs should be an optional argument of the function). This is likely because `MjVFS` doesn't have copy/move constructors, which are required for types used with `std::optional`, causing issues with memory management when std::optional copies the MjVFS object.
To solve this and to make vfs an optional argument to `mj_compile`, I replaced `mj_compile_wrapper` with two overloaded functions. This avoids using `std::optional<MjVFS>` and resolves the memory issue, while still allowing mj_compile to be called with or without the vfs argument from JavaScript.
But what if we instead add those copy/move constructors so it works with `std::optional`?:
While making `MjVFS` copyable or movable would theoretically allow it to work with std::optional, it's not a feasible approach in this case.
The MjVFS struct manages file data in memory, allocating buffers when files are added and freeing them when `mj_deleteVFS` is called in its destructor. A copy would require a deep copy of the entire virtual filesystem, but MuJoCo does not provide a public API function (like a "mj_copyVFS") to do this. Implementing a deep copy manually would be complex and fragile. A shallow copy would result in multiple MjVFS objects pointing to the same memory, leading to double-free errors when their destructors run.
Given these constraints, MjVFS cannot safely be made copyable. The approach of providing two overloaded C++ functions is better.
PiperOrigin-RevId: 855740541
Change-Id: Ia08b6db8868ad245f07145db3fb0322cb9f737cb
`tuple` is better than `list` for constants if we don't expect they get modified by other parts of the code
PiperOrigin-RevId: 838748519
Change-Id: Ic950a78be7ad7518361c046d7e4542607070d3a2
This change introduces a `wrapper_struct_name` function in `structs.py` to consistently generate the name of the C++ wrapper struct from a C struct name. All call sites in `functions.py` and `structs.py` that previously used `common.capitalize` for this purpose now call `wrapper_struct_name`.
PiperOrigin-RevId: 837544242
Change-Id: Ibe5fea8294283dae63bb8d064fd8beb8cfd3f2fe
* All tests moved into wasm/codegen/tests folder
* Merged wasm/codegen/helpers/ into wasm/codegen/generators/
PiperOrigin-RevId: 829471173
Change-Id: I2dbc5d9351771817ec260c7ddf87c31e66a9eabe
Alessio