Early Return Pipeline

[kylebd99]

This PR adds:

1. a no_compute option to LogicExecutor where it returns whatever its interior ctx produces
2. a NoOp stage which immediately returns its inputs

An early return pipeline can then look like, e.g. :

LogicNormalizer(
            LogicExecutor( GalleyLogicalOptimizer(DefaultLoopOrderer(NoOpStage()
                    )
                )

[kylebd99]

Looking at benchmarks/test_galley.py, this PR solves half of the problem. It allows for early return, but it is still a bit inherently complicated to run the first few pipeline stages, intercept the middle, and then benchmark only the remainder. This is because LogicExecutor does things both before and after it calls its child ctx.

[codspeed-hq]

Merging this PR will degrade performance by 64.71%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 1 improved benchmark
❌ 6 regressed benchmarks
✅ 25 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Memory	test_galley_matmul_chain[empty_last-optimize]	54.8 KB	2,447 KB	-97.76%
❌	WallTime	test_galley_matmul_chain[empty_last-optimize]	249.5 ms	5,048.6 ms	-95.06%
❌	WallTime	test_ops_binary[interpret_galley-multiply]	786.2 ms	928.5 ms	-15.33%
❌	WallTime	test_ops_binary[interpret_galley-add]	787 ms	926 ms	-15.01%
❌	WallTime	test_ops_reduction[interpret_galley-sum]	419.5 ms	485.6 ms	-13.61%
❌	WallTime	test_ops_binary[interpret_galley-matmul]	58.9 s	65.7 s	-10.48%
⚡	Memory	test_galley_matmul_chain[dense_last-optimize]	2.6 MB	2.4 MB	+10.74%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing kbd-easier-pipeline-return (5216789) with main (64775fa)}