[Perf] Move attention update stream out of loop to optimize performance

[momo609]

What this PR does / why we need it?

In the update_*attn_params functions, the torch.npu.stream(update_stream) context manager was previously located inside the for-loop that updates parameters for each layer. This resulted in redundant stream initiations for every layer, adding unnecessary overhead.

This commit refactors the code by moving the stream context manager to wrap the entire for-loop. This ensures that the update stream is initiated only once per function call, rather than for each layer. This change reduces 90us in each decode model.
update stream in every layer:

remove update stream in every layer:

Does this PR introduce any user-facing change?

How was this patch tested?

• vLLM version: v0.11.0
• vLLM main: vllm-project/vllm@83f478b

[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 optimizes the update_attn_params function in vllm_ascend/compilation/acl_graph.py by moving the torch.npu.stream context manager outside the for loop. This change effectively reduces the overhead of stream context switching, leading to a performance improvement. The implementation is correct. For consistency and further optimization, consider applying the same pattern to update_mla_attn_params, update_attn_dcp_pcp_params, and update_mla_attn_dcp_pcp_params in the same file, as they exhibit a similar structure.

[whx-sjtu]

LGTM

[yiz-liu]

Please also refactor update_mla_attn_dcp_pcp_params and update_attn_dcp_pcp_params.

[momo609]

Please also refactor update_mla_attn_dcp_pcp_params and update_attn_dcp_pcp_params.
dcp，pcp cannot test now, will add in other pr.

[github-actions]

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