[Refactor] AttentionBuilder inherit from base class in vllm

[LICO1314]

What this PR does / why we need it?

This PR makes AscendMLAMetadataBuilder and AscendSFAMetadataBuilder properly inherit from the base class MLACommonMetadataBuilder in vllm by adding super().__init__() calls.

Changes:

• Add super().__init__() call in AscendMLAMetadataBuilder.__init__()
• Add super().__init__() call in AscendSFAMetadataBuilder.__init__()
• Extract ascend_chunked_prefill_workspace_size() to vllm_ascend/attention/utils.py to avoid code duplication
• Override determine_chunked_prefill_workspace_size() to support Ascend-specific 128k tokens workspace size (vs 64k in parent class)
• Update unit tests to mock parent class __init__ for proper isolation

Why we need it:

• Follow proper Python inheritance patterns by calling super().__init__()
• Reduce code duplication by reusing parent class initialization logic
• Better maintainability as parent class changes will be automatically inherited

Part of issue #5463 item 10

Does this PR introduce any user-facing change?

No, this is an internal refactoring that does not change any user-facing behavior.

[github-actions]

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

• A PR should do only one thing, smaller PRs enable faster reviews.
• Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
• Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.

[gemini-code-assist]

Code Review

This pull request correctly refactors AscendMLAMetadataBuilder and AscendSFAMetadataBuilder to inherit from MLACommonMetadataBuilder, which improves code structure and reduces duplication by leveraging the base class's __init__. The override of determine_chunked_prefill_workspace_size to support Ascend-specific requirements is also a good addition. I have one suggestion to further improve maintainability by avoiding code duplication for this new static method across the two builder classes.

[gemini-code-assist]

This static method determine_chunked_prefill_workspace_size is also duplicated in vllm_ascend/attention/sfa_v1.py. To improve maintainability and adhere to the DRY (Don't Repeat Yourself) principle, consider extracting this method into a common mixin class that both AscendMLAMetadataBuilder and AscendSFAMetadataBuilder can inherit from.

For example, you could create a mixin like this:

class AscendChunkedWorkspaceMixin:
    @staticmethod
    def determine_chunked_prefill_workspace_size(vllm_config: VllmConfig) -> int:
        """Provide workspace size calculation for Ascend NPUs."""
        scheduler_config = vllm_config.scheduler_config
        cache_config = vllm_config.cache_config
        model_config = vllm_config.model_config

        chunked_prefill_workspace_size = min(
            max(8 * model_config.max_model_len,
                4 * scheduler_config.max_num_seqs * cache_config.block_size),
            128 * 1024)

        # Enforce that we have enough for at least 1 page per request
        chunked_prefill_workspace_size = max(
            chunked_prefill_workspace_size,
            scheduler_config.max_num_seqs * cache_config.block_size,
        )

        return chunked_prefill_workspace_size

# Then apply it to the builder classes:
class AscendMLAMetadataBuilder(AscendChunkedWorkspaceMixin, MLACommonMetadataBuilder[AscendMLAMetadata]):
    ...

# in sfa_v1.py
class AscendSFAMetadataBuilder(AscendChunkedWorkspaceMixin, MLACommonMetadataBuilder[AscendSFAMetadata]):
    ...

This would centralize the logic and make future changes easier.

[github-actions]

This pull request has conflicts, please resolve those before we can evaluate the pull request.