[Feat] Export shuffle xor/up/down/idx ops

[sjfeng1999]

Motivation

Technical Details

Test Plan

Test Result

Submission Checklist

• Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

[Copilot]

Pull request overview

Adds module-level GPU warp/subgroup shuffle helpers to the FlyDSL expression layer so kernels can perform lane-crossing data movement (xor/up/down/idx) via the MLIR gpu.shuffle op.

Changes:

• Introduces a shuffle(value, offset, width, mode=...) wrapper around gpu.ShuffleOp, with result re-wrapping via dsl_math_wrap_result.
• Exports convenience entry points shuffle_xor, shuffle_up, shuffle_down, shuffle_idx via __all__.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.