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.
needhull_ was only set while iterating geoms, so a mesh whose inertia is
computed from its convex hull never got one unless some geom pointed at
it. mjCMesh::ComputeVolume and ComputeInertia then read graph_[1] and
GraphFaces() off a null pointer and the compiler crashed.
Whether a mesh needs its hull for inertia is a property of the mesh, not
of any geom, so move that condition out of the geom loop and apply it to
every mesh.
Fixes#3431
Convert the four remaining mju_mulMat{,T}{Vec,Mat}3 call sites in this
file to their inlined mji_ equivalents, so the file is consistent after
the review change to the stretch paths. Aliasing is unaffected: the
inlined variants mark the result restrict, and every call already
required a distinct result (the public versions overwrite res while
reading vec).
Per review on #3437: add the missing blank line above the new comment
block, and use the internal inlined mji_mulMat{,T}{Vec,Mat}3 helpers
rather than the public mju_ ones, matching fe9dc584. Six call sites in
the two stretch paths; pre-existing mju_ calls elsewhere in the file are
left alone.
The implicit effective metric assembles the stretch stiffness from
world-space edge vectors, but a flex vertex body's slide dofs are
expressed in its parent body's frame. When that frame is rotated the
assembled operator is therefore not the Jacobian of the passive stretch
force, which mj_flexPassiveStretch already maps into the dof frame with
xmat^T. The metric is then inconsistent with the force it linearizes:
implicit integration loses its stability guarantee, and models that the
same flex handles comfortably in an unrotated frame diverge.
Apply the matching change of basis in both places that build or apply the
stretch stiffness: mjd_flexStretch_mul rotates the input dof vector into
world and the scattered result back, and mjd_flexStiff_assemble sandwiches
each 3x3 block as R_bi^T * blk * R_bj. Both are no-ops when the parent is
unrotated. Bending needs no change: its blocks are isotropic, and
R^T (q I) R = q I.
This completes the fix in fe9dc584, which covered the passive force paths
and the interp (trilinear) derivative but not the standard stretch one.
On a mesh flex inside a body with a 90-degree rotation, the metric's
directional agreement with the force Jacobian goes from cos = 0.57 to
cos = 1.0, and a hanging sheet that previously reached 176% strain
settles at 0.87%.
This change introduces a get_frame method to the viewer protocol, replacing the previous is_running check. This allows the viewer loop to skip rendering and lifecycle events when no active frame is ready (for example, when no browser is connected to the web viewer).
Additionally, this change:
- Adds a shutdown hook to ViewerHandle to wait for the viewer to cleanly release resources before exiting.
- Implements an ExitEvent handler in the viewer protocol to stop the run loop.
- Adds an http_port configuration option for the web viewer.
- Prevents double-closing in the viewer.
PiperOrigin-RevId: 954603868
Change-Id: Iaba7d131dc94191de51150464514dd10fa7834b0
Avoid mutating ImGui draw lists in place by scaling clip rects locally during scissor computation instead of calling ScaleClipRects. This prevents clip rects from growing exponentially on persistent draw lists, such as streamed remote UI. Additionally, clamp the scaled clip rects to the viewport boundaries to prevent Filament's scissor test from rejecting out-of-bounds rects, while avoiding resetting to full-window scissors which would disable clipping.
The draw geometry is now immutable in the Update function (ImGui texture state is still mutated).
PiperOrigin-RevId: 954277141
Change-Id: Ifdf6866bb4eabaca44b020ea40acc246428b4f0b
Also updates MuJoCo files needed to address changes in Filament.
- "build with clang on windows" patch to sync context lines, remove some
upstream conflicts, disable Clang thread-safety attribute errors, and
allow running on MSYS2.
- Skip building Filament samples.
- Add missing <string_view> to material_manager.cc.
This forces git to treat the current directory as the git checkout to apply the
patch to, which avoids issues where a non-git checkout (e.g. from an unpacked
cached tarball) can cause git to find a higher-level git checkout and try to
apply the patches there, which then fails when the files to patch aren't found.
PiperOrigin-RevId: 953377327
Change-Id: I86f51d4e2823b64c825070139479089c7f30e53e
Fast path in mj_flexPassiveInterp, mj_flexPassiveBendInterp, and
mj_flexPassiveStretch assumed body slide joints are world-aligned
(J = I). When parent body has a non-identity quaternion, joint axes
are rotated (J = R_body), causing wrong force mapping and instability
(NaN/Inf in QACC).
Fix: project world-frame forces onto body local frame via
mju_mulMatTVec3(R_body^T, force) before adding to qfrc_spring/damper.
Also fix the derivative paths:
- mjd_flexInterp_kernel fast path: R^T * K_rot * R * vec
- mjd_flexStiff_assemble (CSR): R_bi^T * K_rot_block * R_bj
The CSR-assembled stiffness matrix is the actual path used by the
implicit CG solver (mj_flexCG gate). The test uses solver="CG" to
activate this path; without it, flex stiffness is integrated
explicitly and no derivative fix can help.
Ported from GitHub PR https://github.com/google-deepmind/mujoco/pull/3379
Original author: Devansh (https://github.com/devansh0703)
Fixes https://github.com/google-deepmind/mujoco/issues/3364
PiperOrigin-RevId: 952789284
Change-Id: If924f7160dd16c0cc88170d80e6e605da0fe2e04
This var was missing from the "clean up" step for all the path vars, causing
the cmake cache lookups to silently fail due to path mismatch.
PiperOrigin-RevId: 952410036
Change-Id: I5c542e821317497e899de95e5c62d76f88836e60
* Enhances ImGui dialogs to support keyboard shortcuts (Enter/Escape) and filters.
* Adjusts GUI layouts for simulation speed and compiler settings.
* Exposes additional model fields in the XML spec generator.
PiperOrigin-RevId: 951988061
Change-Id: I4927c12a8882ba4e91e6098355ec2f5db35c2747
Each pybind extension module holds its own copy of the ImGui/ImPlot globals (extension modules are loaded RTLD_LOCAL), so the viewer that owns the contexts must share the pointers explicitly. Without this, client plotting GUIs and user plotting scripts (e.g. the implot sample) crash on null context pointers.
PiperOrigin-RevId: 951917287
Change-Id: I639a173c2ae9f1af4feaf57aff7b5c81854b7393
Updates CONTRIBUTING.md to suggest that contributors who do not require Git authorship attribution let maintainers know in the PR, which allows for faster direct incorporation of changes.
PiperOrigin-RevId: 951751097
Change-Id: I521f1ad1978987c71a90e85c1bed71392a07ee7b
https://youtu.be/17XpwnqyCXs
New transmission type mjTRN_SO3: a relative orientation, targeting a ball
joint or a site+refsite pair. It is the first transmission with more than
one force output: its length is the norm of the expmap vector of the
relative rotation and its moment axes are the 3 rows of the
relative rotational Jacobian, without projecting onto per-actuator gears.
New force law mjGAIN_SO3/mjBIAS_SO3: a geodesic PD servo, force =
kp * log(q_current^-1 * q_target) - kv * velocity, exact for arbitrary axis
combinations with a unique equilibrium at every commanded orientation.
Error, moment rows and velocity all live in the child frame (joint or
site): the right-difference error is the gradient of the geodesic
potential in that frame. The parent-frame (left) error is not: driving
child-frame torques with it pumps energy at large angles, settling into
steady-spinning limit cycles (the SO3LargeAngleConvergence test). The
integrator variant stores the 3D orientation setpoint in act (actnum = 3,
re-anchored to a bounded representative at integration time). Exposed in
MJCF as <orientation joint=|site=+refsite= kp kv|dampratio>, or via
<general gaintype="so3" biastype="so3">.
The setpoint input has two charts: an expmap target (3 controls, default)
or a quaternion target (4 controls) -- <orientation input="quat">, the
first actuator with different input and output widths. The signature is
recorded in a new per-actuator field actuator_ctrlspec (mjtCtrlChart),
whose meaning is scoped by the gain type the way gain/bias parameters are;
ctrlnum is derived from it at compile time and remains the layout
authority. An explicit field rather than width inference or a prm slot:
width-as-chart cannot express same-width signatures (upcoming servo input
subsets), and prm slots are the input_mode pattern this stack retires.
The force law normalizes the commanded quaternion, making it scale- and
antipodally-invariant. The all-zero ctrl still maps to the identity via
mju_normalize4, but it is a degenerate point (a nudge of any component
commands a half-turn), so quat inputs reset to the identity quaternion:
new mj_resetCtrl sets neutral ctrl values (zero, except qw = 1), called
by mj_resetData and the viewers' Clear All. The quat chart is
restricted to dyntype 'none': integrating a quaternion setpoint linearly
is not meaningful on the manifold. New mjsActuator.ctrlspec field carries
the signature through the spec and XML round-trip.
Actuator sensors (actuatorpos/vel/frc) now report one value per force
output; dim = 3 on an SO3 actuator.
As the first actuator with nu != nactuator, this commit also makes the
viewers multi-input aware: the control sliders in simulate and studio,
which indexed per-actuator arrays by control index (out of bounds on
this model class), are generated per control and labeled with the
actuator name plus an input suffix ("orient/qw"), via the new
introspection helper mj_actuatorInputName -- the single source of truth
for input names, extended by each new multi-input type (quaternion
components are w-first: qw, qx, qy, qz). Slider ranges now honor a
defined ctrlrange even when ctrllimited is false: range is the UI hint,
limited is the clamp -- wrapped and expmap setpoints are unbounded but
still want finite sliders, while quat components are truly bounded.
The rotational demo model is orientation.xml under
test/engine/testdata/actuation/, upgraded to a three-way contrast:
per-axis wrapped servos vs an expmap-commanded vs a quat-commanded
orientation actuator, on identical checker-textured boxes. It is loaded
by the mixed-axis contrast and input-name tests, and doubles as the
viewer test model (slider groups of 3 independent, 3 grouped, 4 grouped).
PiperOrigin-RevId: 951607063
Change-Id: If235dba8e2f2ca72672e7c62531a27e967c6a373
RenderableManager is responsible for rendering the "skinned" flexes
which represent the correct "final" look of a flex. This now includes
1D flexes which are rendered as a sequence of capsules (similar to
tendons).
SceneDecorator is responsible for rendering the debug (unskinned)
versions of flexes. This includes the vertices (small spheres for each
vertex), edges (thin cynlinders for each edge), and faces (triangles
for each face, including sublayers).
PiperOrigin-RevId: 951595046
Change-Id: Iebc6fd7a747a602813b6457333bd9fa5915e28d3
Yuval Tassa