[Bugfix] Fix FusedMoEPrepareAndFinalize for cuda-disalike backends

[MengqingCao]

Set FusedMoEPrepareAndFinalize to placeholder None for cuda-disalike backends, cause it is using everywhere in vllm/model_executor/layers/fused_moe/layer.py as a type indicator

[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.

🚀

[MengqingCao]

cc @DarkLight1337 @youkaichao This bug is introduced by #15956, I think we should merge this fix asap, thx!

[MengqingCao]

@DarkLight1337 CI failed due to pp process group initialization issue in spec-decode, seems no relation to this pr, could we merge this first?