[CB] remove env var VLLM_SPYRE_ENABLE_PREFILL_OPTIMIZATION

[yannicks1]

VLLM_SPYRE_ENABLE_PREFILL_OPTIMIZATION is on by default and we have not found any reason to ever turn it off.

reasons for removing:

• getting rid of dead code -> simplifying scheduler constraints
• being consistent with chunked prefill scheduler: the variable is not used there
• reducing GHA time by removing tests targeting this optimization specifically

[github-actions]

👋 Hi! Thank you for contributing to vLLM support on Spyre.
Just a reminder: Make sure that your code passes all the linting checks, otherwise your PR won't be able to be merged. To do so, first install the linting requirements, then run format.sh and commit the changes. This can be done with uv directly:

uv sync --frozen --group lint --active --inexact

Or this can be done with pip:

uv pip compile --group lint > requirements-lint.txt
pip install -r requirements-lint.txt
bash format.sh

Now you are good to go 🚀

[maxdebayser]

Thanks, this makes sense.

[sducouedic]

there are still two tests with referring to prefill optimization:

test_requests_use_full_batch_tkv_limit_prefill_opt and test_requests_exceed_batch_tkv_limit_prefill_opt

[yannicks1]

good catch, will change the names of these tests.

[sducouedic]

LGTM, definitively a no brainer to have the optimization enabled all the time