[Tile] Mark the thrust actor framework as tile enabled.

[miscco]

Those are purely forwarding structs, so they should be able to take whatever is thrown at them

[coderabbitai]

📝 Walkthrough

Summary by CodeRabbit

• Bug Fixes
  • Improved compatibility and diagnostics for public functional utilities and type-erased assignment operations.
  • Preserved existing runtime behavior while refining how these APIs are exposed and checked across supported execution environments.

Walkthrough

Changes

The pull request replaces _CCCL_HOST_DEVICE annotations with _CCCL_API and, where applicable, _CCCL_EXEC_CHECK_DISABLE on functional actor APIs and any_assign operations. Runtime behavior remains unchanged.

Annotation updates

Layer / File(s)	Summary
Functional API annotation migration thrust/thrust/detail/functional/actor.h, thrust/thrust/iterator/detail/any_assign.h	Updated annotations on actor, argument, composite, operator adaptor, value, composition helpers, and any_assign constructors and assignment operators. Existing evaluation, forwarding, construction, and assignment behavior remains unchanged.

Suggested reviewers: elstehle

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

🧹 Nitpick comments (2)

thrust/thrust/detail/functional/actor.h (2)

39-40: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick win

suggestion: Confirm the annotation policy for defaulted constructors.

These declarations add _CCCL_EXEC_CHECK_DISABLE, but they do not add _CCCL_API. _CCCL_EXEC_CHECK_DISABLE suppresses an execution-space check. It does not provide the function API annotation. (nvidia.github.io)

If explicitly defaulted special members are intentionally exempt because CUDA infers their execution space, document that exemption and cover host, device, and tile construction in a compile test. (docs.nvidia.com)

As per coding guidelines, “Functions must be marked with _CCCL_HOST_API, _CCCL_DEVICE_API, _CCCL_HOST_DEVICE_API, _CCCL_TILE_API, or _CCCL_API.” As per path instructions, preserve Thrust API compatibility and host/device callability.

Also applies to: 96-97, 122-123, 156-157

Sources: Coding guidelines, Path instructions, MCP tools

---

42-56: 🎯 Functional Correctness | 🔵 Trivial

suggestion: Run a focused host/device/tile compile test.

Use the relevant CCCL CMake preset and Thrust target. Test normal SIMT and tile-enabled compilation. Exercise actor construction, operator(), eval, make_actor, both compose overloads, and any_assign with arbitrary input types. Verify dispatch and ADL behavior.

No targeted build or test result is included in the provided context.

As per coding guidelines, “Validate changes with targeted builds and tests.” As per path instructions, verify API compatibility, dispatch behavior, host/device callability, and templated arbitrary-input support.

#!/usr/bin/env bash
set -euo pipefail

rg -n 'make_actor|compose\(|any_assign|_CCCL_API|_CCCL_EXEC_CHECK_DISABLE' . \
  --glob '*.cpp' --glob '*.cc' --glob '*.cxx' --glob '*.cu' \
  --glob '*.h' --glob '*.hh' --glob '*.hpp' --glob '*.cuh'

cmake --list-presets=all

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: bb985407-ffdb-416b-b482-2f08725f84bd

📥 Commits

Reviewing files that changed from the base of the PR and between 2bff086 and 055cd27.

📒 Files selected for processing (2)

• thrust/thrust/detail/functional/actor.h
• thrust/thrust/iterator/detail/any_assign.h

[github-actions]

⏱️ CCCL compile-time benchmark comparison: Public headers compile-time bench

Result: 0 regression row(s), 6 improvement row(s) above threshold.

Run	Value
Config	public-headers-gcc13
Baseline	origin/main
Preset	all-dev
Targets	cub.headers.base, thrust.cpp.cuda.headers.base, libcudacxx.test.public_headers
GPU / launch args	rtx2080 / --cuda 13.3 --host gcc13

Artifacts: reports and traces

Direct file processing

-f file-processing exclusive --sort total

🟢 Direct file processing — Improvements

Rank	Improvement impact	Selected Δ	Baseline	Current	Event	Matched traces
1	0.752937	-0.752937	9.336447	8.583510	Processing Header File: libcudacxx/include/cuda/std/__cccl/prologue.h	550
2	0.472273	-0.472273	3.331492	2.859219	Processing Header File: libcudacxx/include/cuda/std/__cccl/epilogue.h	550
3	0.439944	-0.439944	7.982951	7.543007	Processing Header File: cub/cub/util_type.cuh	279
4	0.382773	-0.382773	6.315579	5.932806	Processing Header File: libcudacxx/include/cuda/std/__tuple_dir/vector_types.h	238
5	0.257114	-0.257114	4.126131	3.869017	Processing Header File: libcudacxx/include/cuda/std/__utility/pair.h	404
6	0.210818	-0.210818	20.701856	20.491038	Processing Header File: libcudacxx/include/cuda/__device/physical_device.h	94