Skip to content

Conversation

@ingowald
Copy link
Contributor

added (to rayQueries.h) new routines

  • cuBQL::shrinkingRayQuery::twoLevel::forEachLeaf()
    and
  • cuBQL::shrinkingRayQuery::twoLevel::forEachPrim()
    In addition to the regular BVH traversals these variants have enterBlas() and leaveBlas() lambdas that allow the user to provide a transformed ray and blas BVH when toplevel BVH reaches a leaf. Managing transforms are the job ob the user, but these lambdas allow for doing so.

- added 'twoLevel::' namespace (and corresponding abstraction) to
  shrinking radius query. Twolevle traversal has addtl 'enterBlas()'
  and 'leaveBlas()' lambdas that can modify ray and bvh node ptr
  when reaching tlas leaves
  twoLevel variants get additional enterBlas() and leaveBlas() lambdas so
  traversal can traverse two nested BVHes at the same time.
- unified Triangle3f and triangle_t<>; former is now just
  triangle_t<float>. Same for TriangleIntersection and
  TriangleIntersection::compute()
- various smaller clanups and warning fixes
@iwald-nvidia iwald-nvidia merged commit 52294ac into NVIDIA:main Dec 15, 2025
10 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants