[SandroMaglione]

Summary

• Since 0.45.0, a child's worker starts in the turn of the invocation that completes. If the child's onDone sends to its parent and the parent's transition stops that child, the stop asked to interrupt the worker from inside the invocation fiber. The invocation fiber then waited for the child's worker, and the worker was waiting for the invocation fiber to exit, so both machines hung, including scope finalization.
• startWorker now forks the requested interruption instead of awaiting it in the sender's turn. Whoever requested the stop already awaits completion.
• Adds a regression test covering synchronous and asynchronous invocations, with generic and compiled parents. The asynchronous cases hang on main.

Changeset

• Added or updated for a library or package-metadata change
• Not required because this PR does not change a publishable package

Validation

• pnpm check
• Automated type-performance measurement passed or was not required
• Automated runtime- and memory-performance measurement passed or was not required

pnpm perf:runtime also completed locally. A consumer app's machine tests that hung on 0.45.0 pass with this build: 357/357.

🤖 Generated with Claude Code

[github-actions]

Type performance

Measured with TypeScript 6.0.3 and skipLibCheck=true.

Scenario	Base	PR	Difference
Effect only	55	55	0 (0.0%)
Import effect-machine	55	55	0 (0.0%)
Machine.state (3 children)	2,664	2,664	0 (0.0%)
Machine.make (3 states, 2 events)	8,932	8,932	0 (0.0%)
machine.handle (3 states, 2 transitions)	39,449	39,449	0 (0.0%)
declared transition (10 named branches)	60,402	60,402	0 (0.0%)
atomic construction, guards, and reentry	85,710	85,710	0 (0.0%)
registered invocation (input-taking Effect)	39,819	39,819	0 (0.0%)
machine.handle (depth 24)	194,531	194,531	0 (0.0%)
machine.handle (wide depth 16)	218,238	218,238	0 (0.0%)
machine.handle (parallel/history/choice)	91,642	91,642	0 (0.0%)
machine definition (3 independent implementations)	99,173	99,173	0 (0.0%)
machine exact input/output/error/services	82,941	82,941	0 (0.0%)
execution adapter readiness	71,428	71,428	0 (0.0%)

Marginal instantiations are measured against the matching setup without that API call:

Scenario	Base	PR	Difference
Import effect-machine	0	0	0
Machine.state (3 children)	2,609	2,609	0 (0.0%)
Machine.make (3 states, 2 events)	6,260	6,260	0 (0.0%)
machine.handle (3 states, 2 transitions)	30,517	30,517	0 (0.0%)
declared transition (10 named branches)	51,849	51,849	0 (0.0%)
atomic construction, guards, and reentry	70,370	70,370	0 (0.0%)
registered invocation (input-taking Effect)	32,682	32,682	0 (0.0%)
machine.handle (depth 24)	177,222	177,222	0 (0.0%)
machine.handle (wide depth 16)	201,707	201,707	0 (0.0%)
machine.handle (parallel/history/choice)	79,144	79,144	0 (0.0%)
machine definition (3 independent implementations)	88,492	88,492	0 (0.0%)
machine exact input/output/error/services	70,732	70,732	0 (0.0%)
execution adapter readiness	31,922	31,922	0 (0.0%)

Check times (informational)

Scenario	Base	PR
Effect only	0.02s	0.02s
Import effect-machine	0.02s	0.02s
Machine.state (3 children)	0.05s	0.06s
Machine.make (3 states, 2 events)	0.08s	0.09s
machine.handle (3 states, 2 transitions)	0.19s	0.19s
declared transition (10 named branches)	0.23s	0.23s
atomic construction, guards, and reentry	0.31s	0.32s
registered invocation (input-taking Effect)	0.21s	0.19s
machine.handle (depth 24)	0.39s	0.37s
machine.handle (wide depth 16)	0.42s	0.38s
machine.handle (parallel/history/choice)	0.35s	0.32s
machine definition (3 independent implementations)	0.34s	0.33s
machine exact input/output/error/services	0.33s	0.33s
execution adapter readiness	0.28s	0.27s

Type instantiations are the comparison metric. Check time varies with runner load and is informational only.

[github-actions]

Runtime performance

Median of 5 independent benchmark processes on AMD EPYC 9V45 96-Core Processor with Node v24.21.0.

Pull request baseline

Scenario	Effect Machine
Plan counter transitions	187,719 transitions/s
Drain burst with terminal fence	663,024 increments/s
Drain burst with a change observer	625,433 increments/s
Lookup and send to one child	545,340 increments/s
Start and stop a machine	198,926 machines/s
Start and stop a parent with one child	61,675 families/s
Complete an Effect invocation with tracing enabled	26,756 invocations/s
Complete an Effect invocation with tracing disabled	29,481 invocations/s
Plan transitions through a compound state	170,125 transitions/s
Plan transitions through parallel regions	136,273 transitions/s
Drain burst through a compound state	546,437 events/s
Drain burst through two parallel regions	511,291 events/s
Drain a compound-state burst with a change observer	518,622 events/s

Process runtime reference points

Scenario	Effect Machine
Start and stop a raw generic process	53,425 processes/s
Start and stop a raw compiled process	121,921 processes/s

Memory profile	Effect Machine
Idle machine	1.9 KiB
Raw generic managed process	11.3 KiB
Raw compiled process	3.1 KiB
Two independent idle machines	3.5 KiB
Idle parent with one child	6.0 KiB
Parent with observed child registry	9.5 KiB
Parent with observed invoked child snapshots	6.9 KiB

Effect Machine change from base

Metric	Base	Base variability	PR	PR variability	Difference
Plan counter transitions	192,567 transitions/s	0.1% MAD	187,719 transitions/s	0.6% MAD	-2.5%
Drain burst with terminal fence	681,598 increments/s	0.5% MAD	663,024 increments/s	0.6% MAD	-2.7%
Drain burst with a change observer	628,850 increments/s	0.7% MAD	625,433 increments/s	0.9% MAD	-0.5%
Lookup and send to one child	556,368 increments/s	1.0% MAD	545,340 increments/s	2.0% MAD	-2.0%
Start and stop a machine	198,886 machines/s	0.4% MAD	198,926 machines/s	0.8% MAD	+0.0%
Start and stop a parent with one child	61,828 families/s	0.4% MAD	61,675 families/s	0.1% MAD	-0.2%
Complete an Effect invocation with tracing enabled	26,784 invocations/s	2.1% MAD	26,756 invocations/s	1.5% MAD	-0.1%
Complete an Effect invocation with tracing disabled	30,112 invocations/s	1.0% MAD	29,481 invocations/s	0.9% MAD	-2.1%
Plan transitions through a compound state	170,328 transitions/s	0.1% MAD	170,125 transitions/s	1.1% MAD	-0.1%
Plan transitions through parallel regions	136,766 transitions/s	1.2% MAD	136,273 transitions/s	1.1% MAD	-0.4%
Drain burst through a compound state	550,509 events/s	0.5% MAD	546,437 events/s	0.4% MAD	-0.7%
Drain burst through two parallel regions	506,994 events/s	0.5% MAD	511,291 events/s	0.9% MAD	+0.8%
Drain a compound-state burst with a change observer	517,848 events/s	0.6% MAD	518,622 events/s	1.0% MAD	+0.1%
Idle machine heap per unit	1.9 KiB	1.6% MAD	1.9 KiB	1.3% MAD	+2.2%
Raw generic managed process heap per unit	11.2 KiB	0.0% MAD	11.3 KiB	1.4% MAD	+0.7%
Raw compiled process heap per unit	3.1 KiB	0.1% MAD	3.1 KiB	0.1% MAD	+0.1%
Two independent idle machines heap per unit	3.5 KiB	0.0% MAD	3.5 KiB	0.1% MAD	-0.6%
Idle parent with one child heap per unit	6.0 KiB	0.1% MAD	6.0 KiB	0.0% MAD	-0.0%
Parent with observed child registry heap per unit	9.5 KiB	0.0% MAD	9.5 KiB	0.0% MAD	+0.0%
Parent with observed invoked child snapshots heap per unit	6.9 KiB	0.1% MAD	6.9 KiB	0.2% MAD	+0.3%

Process runtime reference change from base

Metric	Base	Base variability	PR	PR variability	Difference
Start and stop a raw generic process	52,554 processes/s	1.1% MAD	53,425 processes/s	1.2% MAD	+1.7%
Start and stop a raw compiled process	120,019 processes/s	0.5% MAD	121,921 processes/s	0.1% MAD	+1.6%

Regression guard

No large, noise-adjusted throughput or heap regressions detected.

Versions and interpretation

• Effect Machine: 0.45.0

Higher throughput is better; lower heap is better. Variability is the median absolute deviation across independent processes, relative to their median. Small differences on shared GitHub-hosted hardware remain informational; the required guard rejects only large changes beyond the measured noise allowance.