Fix cuda::ptx shr and bfind long, long long handling

[fbusato]

Description

Fix cuda::ptx shr and bfind long, long long (+ unsigned) handling across Linux and Windows

Fixes #9992

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 600c17f8-19fc-430a-8fb0-64b4b052b2d4

📥 Commits

Reviewing files that changed from the base of the PR and between b757d02 and dba1bf7.

⛔ Files ignored due to path filters (6)

• docs/libcudacxx/ptx/instructions/generated/bfind.rst is excluded by !**/generated/**
• docs/libcudacxx/ptx/instructions/generated/shr.rst is excluded by !**/generated/**
• libcudacxx/include/cuda/__ptx/instructions/generated/bfind.h is excluded by !**/generated/**
• libcudacxx/include/cuda/__ptx/instructions/generated/shr.h is excluded by !**/generated/**
• libcudacxx/test/libcudacxx/cuda/ptx/generated/bfind.h is excluded by !**/generated/**
• libcudacxx/test/libcudacxx/cuda/ptx/generated/shr.h is excluded by !**/generated/**

📒 Files selected for processing (2)

• libcudacxx/include/cuda/__ptx/instructions/bfind.h
• libcudacxx/include/cuda/__ptx/instructions/shr.h

---

📝 Walkthrough

Summary by CodeRabbit

• Bug Fixes
  • Improved compatibility and reliability for CUDA PTX bit-manipulation instruction support across integral and signed data types.

Walkthrough

Changes

PTX instruction type-trait dependencies

Layer / File(s)	Summary
Add type-trait includes libcudacxx/include/cuda/__ptx/instructions/bfind.h, libcudacxx/include/cuda/__ptx/instructions/shr.h	The PTX instruction headers now include CUDA standard-library traits for integral and signedness checks.

Assessment against linked issues

Objective	Addressed	Explanation
Ensure all signed integer types use arithmetic shift in cuda::ptx::shr [#9992]	❓	The required traits are included, but no overload or dispatch logic changes are shown to establish arithmetic shifting for long long or other signed integer types.

Out-of-scope changes

Code Change	Explanation
Add is_integral, is_signed, and is_unsigned includes to bfind.h (libcudacxx/include/cuda/__ptx/instructions/bfind.h:27-29)	The linked issue concerns cuda::ptx::shr; no stated objective covers bfind dependencies.

Possibly related PRs

• NVIDIA/cccl#9993: Updates the PTX shr header with type-trait includes related to signed-shift dispatch.

---

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

[github-actions]

😬 CI Workflow Results

🟥 Finished in 2h 42m: Pass: 99%/120 | Total: 4d 06h | Max: 2h 35m | Hits: 44%/1345151

See results here.