Add vLLM-metax vllm metax cmake presets jobs override

[ghangz]

Summary

• Adds a focused vllm metax cmake presets jobs override improvement for MetaX-MACA/vLLM-metax.
• The change targets MetaX MACA development and validation workflows, with emphasis on earlier diagnostics, reproducible logs, or safer benchmark tooling.
• Existing default behavior is kept compatible; the new logic is scoped to explicit checks, helper tools, or validation metadata.

Validation

• Verified on Gitee.AI MetaX GPU resources: vLLM-metax_vLLM image batch, 10/10 PASS; PyTorch-MACA batch also covered vLLM-metax runtime tools.
• Branch validation command: python -m pytest tests/tools/test_generate_cmake_presets_jobs.py
• Pull request text is intentionally ASCII-only to avoid encoding issues on web forms and API clients.

Review notes

• Source branch: ghangz:mengz/vllm-metax-cmake-presets-jobs-override
• Target branch: MetaX-MACA/vLLM-metax:master
• Maintainers can modify this branch if follow-up adjustments are needed.

[gemini-code-assist]

Code Review

This pull request adds support for overriding the number of CMake build jobs and NVCC threads via command-line arguments and function parameters, and includes a new unit test to verify this behavior. The reviewer identified a potential bug where get_cpu_cores() returning 0 or None could lead to a ZeroDivisionError or TypeError, and provided a robust code suggestion to handle these edge cases.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

If get_cpu_cores() returns None (which multiprocessing.cpu_count() can do on some platforms) or returns 0 (which can happen in certain containerized or restricted environments), the current logic will raise a TypeError or a ZeroDivisionError.

Specifically, if cpu_cores is 0, nvcc_threads is set to 0 via min(4, 0). Then, calculating cmake_jobs using cpu_cores // nvcc_threads results in 0 // 0, which raises a ZeroDivisionError.

We can make this logic robust against both None and 0 values by defaulting cpu_cores to at least 1 and ensuring nvcc_threads is at least 1.

Suggested change

	if nvcc_threads is None:
	nvcc_threads = min(4, cpu_cores)
	elif nvcc_threads < 1:
	raise ValueError("nvcc_threads must be at least 1")
	if cmake_jobs is None:
	cmake_jobs = max(1, cpu_cores // nvcc_threads)
	elif cmake_jobs < 1:
	raise ValueError("cmake_jobs must be at least 1")
	if nvcc_threads is None:
	nvcc_threads = max(1, min(4, cpu_cores or 1))
	elif nvcc_threads < 1:
	raise ValueError("nvcc_threads must be at least 1")
	if cmake_jobs is None:
	cmake_jobs = max(1, (cpu_cores or 1) // nvcc_threads)
	elif cmake_jobs < 1:
	raise ValueError("cmake_jobs must be at least 1")