[Neuron] Add custom_ops for neuron backend

[liangfu]

As part of the effort in supporting vLLM V1 architecture for neuron backend (#11152), this PR intent to support activation, layernorm, rotary_embedding and logits_processor as a variant of existing modules that derive from CustomOp class.

These changes are tested in individual test cases that are ported from tests/kernels directory to be neuron backend specific.

Co-authored-by: George Novack (@gnovack)
Co-authored-by: Aoyu Zhang (@AoyuQC)

[github-actions]

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

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

[mergify]

This pull request has merge conflicts that must be resolved before it can be
merged. Please rebase the PR, @liangfu.

https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork

[gnovack]

Thanks @liangfu! lgtm