[Fix] fix assertion error for chunked prefill when disabling cache

[wangraying]

Motivation

When using chunked prefill with disabled radix cache, the scheduler will fail at the assertion error here.

The error is described in detail in #2255.

Modifications

This PR simply skips chunking the current request if there is already an inflight request, following the same code path as in add_one_req.

Checklist

• Format your code according to the Contributor Guide.
• Add unit tests as outlined in the Contributor Guide.
• Update documentation as needed, including docstrings or example tutorials.

[HaiShaw]

@wangraying thanks for the PR.

[merrymercy]

This is probably not the correct fix. I accepted it accidentally so I reverted it in #2286.
@hnyls2002 will submit a correct fix soon.
Thanks anyway.