Pr/xdna more kernels

[Victor-Jung]

This PR adds support for additional operators (SiLU and LayerNorm) to the XDNA2 platform and refactors aspects of the MLIR code-generation backend to improve scalability. Notably, we make the MLIR code transformation agnostic to kernel metadata, require the use of the tile sizes returned by the tiler, and move the "no remainder tile" constraint to TCF.

Added

• Fp16 LayerNorm and SiLU tests, kernels, binding, and type checking.

Changed

• Update DeeployTest main for XDNA2 to support several inputs.
• Move kernel metadata to the NodeTemplate instead of the CodeTransformation
• Use the tiler as the source of truth for all tile size decisions

PR Merge Checklist

1. The PR is rebased on the latest devel commit and pointing to devel.
2. Your PR reviewed and approved.
3. All checks are passing.
4. The CHANGELOG.md file has been updated.
5. If the docker was modified, change back its link after review.