[Example] Add RoPE backward kernel with mask-based implementation

[liujaaa]

• Implements RoPE forward and backward kernels
• Supports TND [BS, N, D] and BSND [B, S, N, D] input formats
• Supports float16, bfloat16, float dtypes with FP32 accumulation
• Host-side pre-computed mask for gather operation
• Autograd Function wrapper for PyTorch integration

[github-actions]

👋 Hi! Thank you for contributing to the TileLang project.

Please remember to run bash format.sh in the root directory of the project to ensure your changes are properly linted and formatted. This will help ensure your contribution passes the format check.

We appreciate you taking this step! Our team will review your contribution, and we look forward to your awesome work!

🚀