[alxbilger]

This scene crashes in:

• [Core,IntegrationScheme] Replace legacy propagation by mapping-graph based propagation #6334

More specifically because of:

- m_mop->propagateDx(m_r1);
+ simulation::common::DifferentialOperations::pushforwardTangent(m_mappingGraph, *m_params, m_r1);

The first thing we can notice is that DeformableOnRigidFrameMapping has an input belonging to a different solver!

Why is it a problem?

EulerImplicitSolver creates a new Data in each state of its mapping graph. If one state is out of its mapping graph, the Data will not be created. It's the case for m_r1.

Why does it not crash with propagateDx ?

Because in propagateDx, the VecId is defaulted to core::vec_id::write_access::dx when the VecId is not found in the state. So, it does not crash because we access to the Data associated to dx. And it's a problem because we are supposed to access to m_r1...

Why not fixing the scene ?

It can be "fixed" by moving all the objects under the same solver. This is what is done in the similar scene DeformableOnRigidFrameMapping.scn. The scene I deleted is equivalent, but using constraints instead. But for some reasons, the object teleports at the first step. I don't know why...

---

By submitting this pull request, I acknowledge that
I have read, understand, and agree SOFA Developer Certificate of Origin (DCO).

---

Reviewers will merge this pull-request only if

• it builds with SUCCESS for all platforms on the CI.
• it does not generate new warnings.
• it does not generate new unit test failures.
• it does not generate new scene test failures.
• it does not break API compatibility.
• it is more than 1 week old (or has fast-merge label).