Option to Pass Parameters to Scalable CCD Narrow-Phase

[antoinebou12]

Description

This PR fixes GPU CCD (SweepAndTiniestQueue broad phase) ignoring user-specified TightInclusionCCD parameters in compute_collision_free_stepsize.

Previously, the GPU path called scalable_ccd::cuda::ipc_ccd_strategy with hardcoded TightInclusionCCD::DEFAULT_TOLERANCE and TightInclusionCCD::DEFAULT_MAX_ITERATIONS, even when the user passed a custom TightInclusionCCD via the narrow_phase_ccd parameter.

Now, when narrow_phase_ccd is a TightInclusionCCD instance, the GPU path extracts tolerance and max_iterations using dynamic_cast and forwards them to scalable_ccd::cuda::ipc_ccd_strategy. If narrow_phase_ccd is not a TightInclusionCCD instance, we fall back to default values to preserve backward compatibility.

Motivation / Context

• Users can tune CCD precision/performance on CPU via the NarrowPhaseCCD interface, but couldn't do the same for GPU CCD.
• This aligns GPU CCD behavior with the CPU code paths and resolves the parameter-passing gap in issue No Option to Pass Parameters to Scalable CCD Narrow-Phase #124.

Dependencies

• No new dependencies.
• Requires existing CUDA support (IPC_TOOLKIT_WITH_CUDA=ON) and existing TightInclusionCCD class.

Fixes #124

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• Enhancement (non-breaking change which improves existing functionality)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Added a new test case "GPU CCD parameter extraction" in tests/src/tests/ccd/test_gpu_ccd.cpp that verifies:

• Custom TightInclusionCCD parameters (tolerance and max_iterations) are extracted and used on the GPU path
• Default TightInclusionCCD works with default parameters
• Non-TightInclusionCCD types (e.g., AdditiveCCD) fall back to defaults correctly
• Explicit default parameters match implicit defaults, proving parameter extraction is working

Test Instructions:

1. Build with CUDA support enabled (IPC_TOOLKIT_WITH_CUDA=ON)
2. Run: ctest -R test_gpu_ccd
3. Confirm the new "GPU CCD parameter extraction" test case passes

Test Configuration:

• OS and Version: Windows 11 / Linux (WSL)
• Compiler and Version: gcc 10+ / clang (with CUDA 12.6 support)

Checklist

• I have followed the project style guide
• My code follows the clang-format style guidelines of this project
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules

[zfergus]

LGTM. Thanks for the fix!