[TRTLLM-15344][feat] nixl bounce v2 (1/6): foundation components

[chuangz0]

@coderabbitai summary

Description

Part 1/6 of splitting the NIXL bounce-buffer v2 transport (#15780) into reviewable pieces. This PR adds the dependency-free foundation components under cpp/tensorrt_llm/executor/cache_transmission/nixl_utils/bounce/:

• BuddyAllocator — power-of-two buddy allocator used for bounce-buffer credits
• BounceArena — shared GPU data arena (alloc / address mapping / shutdown)
• ExecPool — execution-context pool (stream / event / scratch / host-pinned buffers)
• GatherScatterKernel — CUDA gather/scatter kernel for staging KV blocks
• BounceNvtx — NVTX instrumentation helpers

Nothing is wired into production paths yet; the code is only reachable from its unit tests. The transport core and the NixlTransferAgent integration land in later parts.

Stack: part 1/6 · umbrella: #15780 (full design & motivation)

Test Coverage

Pre-merge CI wiring is included: this PR adds the executor_bounce ctest group (cpp/test_unit_tests.py::test_unit_tests[executor_bounce-90], H100 pre_merge in l0_h100.yml). ctest discovers tests by directory, so the tiers added by later parts of the stack run automatically without further CI changes; part 6 tightens the group with a built-artifact guard for the NIXL E2E test.

• cpp/tests/unit_tests/executor/bounce/buddyAllocatorTest.cpp
• cpp/tests/unit_tests/executor/bounce/bounceArenaTest.cpp (GPU)
• cpp/tests/unit_tests/executor/bounce/execPoolTest.cpp (GPU)
• cpp/tests/unit_tests/executor/bounce/gatherScatterKernelTest.cpp (GPU)

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

🤖 Generated with Claude Code

[chuangz0]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #65810 [ run ] triggered by Bot. Commit: 43a047a Link to invocation

[tensorrt-cicd]

PR_Github #65810 [ run ] completed with state SUCCESS. Commit: 43a047a
/LLM/main/L0_MergeRequest_PR pipeline #53514 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation