Drive PR #2959 (Migrating cuda-python sample codes from cuda-samples repository and Testing) across the finish line #2866

Context

pr #2959 previously 2266 — Migrating cuda-python samples from cuda-samples repository and Testing — has been open since June 25, 2026 and represents significant work (40 commits, +38863/-20564 across 247 files) to establish cuda-python as the canonical source of truth for CUDA samples, which will then be synced back to the cuda-samples repo on each release.

Current Status

• ✅ CI is green (as of Sep 9, 2026)
• ✅ Approved by @rparolin
• ✅ All existing examples from cuda_bindings/examples and cuda_core/examples have been migrated into samples/cuda_bindings and samples/cuda_core
• ✅ Samples are integrated into the main cuda.core and cuda.bindings test suites (no longer a separate suite)
• ✅ Platform-specific waivers added (WSL, Windows TCC, headless OpenGL, managed memory, etc.)
• ⏳ Awaiting final review. @leofang

Remaining Blockers

1. Review from @leofang — the last requested reviewer; PR cannot merge until their review is complete or dismissed
2. Documentation checklist — the "documentation is up to date" checklist item in the PR is unchecked; needs clarification whether it's intentional follow-up work or needs to be addressed now
3. mdboom's "changes requested" status — needs to be resolved/dismissed once all concerns are confirmed addressed

Ask

Please prioritize reviewing and merging PR #2266. The PR is now in a good state with green CI. If it's too large to review in one pass, consider whether splitting into smaller follow-up PRs (as suggested by @lijinf2) is feasible to accelerate the final mile.

Ref: #2266
Milestone: cuda.bindings 13.4.0 & 12.9.8
Labels: P0, cuda.core, cuda.bindings, example