fix(pack): script external export compitable

[fireairforce]

Summary

update from: utooland/next.js#141

Test Plan

updated

[gemini-code-assist]

Code Review

This pull request updates the runtime logic for external modules to use namespace objects and modifies script loading to use an unconditional cross-origin assignment. It also adds jszip as an external dependency and updates the next.js subproject. Feedback suggests that the unconditional cross-origin assignment could lead to incorrect CORS behavior and that for...in loops used for module property copying should be replaced with safer alternatives like Object.keys to avoid prototype pollution.

[github-actions]

📊 Performance Benchmark Report (with-antd)

Utoopack Performance Report

Report ID: utoopack_performance_report_20260427_053515
Generated: 2026-04-27 05:35:15
Trace File: trace_antd.json (0.6GB, 1.62M spans)
Test Project: examples/with-antd

---

Executive Summary

Metric	Value	Assessment
Total Wall Time	11,271.9 ms	Baseline
Total Thread Work (de-duped)	36,052.3 ms	Non-overlapping busy time
Effective Parallelism	3.2x	thread_work / wall_time
Working Threads	5	Threads with actual spans
Thread Utilization	64.0%	🆗 Average
Total Spans	1,624,923	All B/E + X events
Meaningful Spans (>= 10us)	381,165	(23.5% of total)
Tracing Noise (< 10us)	1,243,758	(76.5% of total)

---

Build Phase Timeline

Shows when each build phase is active and how much CPU it consumes.
Self-Time is the time spent exclusively in that phase (excluding children).

Phase	Spans	Inclusive (ms)	Self-Time (ms)	Wall Range (ms)
Resolve	84,053	5,845.3	4,768.0	8,386.6
Parse	10,224	2,827.2	2,321.4	11,136.6
Analyze	224,572	25,211.7	18,238.9	10,886.9
Chunk	22,898	2,202.8	2,056.6	8,713.2
Codegen	30,123	4,433.8	3,278.2	8,796.2
Emit	38	42.9	21.3	8,708.6
Other	9,257	1,417.9	1,140.6	11,271.9

---

Workload Distribution by Diagnostic Tier

Category	Spans	Inclusive (ms)	% Work	Self-Time (ms)	% Self
P0: Scheduling & Resolution	315,308	31,939.4	88.6%	23,629.2	65.5%
P1: I/O & Heavy Tasks	3,051	327.4	0.9%	305.8	0.8%
P2: Architecture (Locks/Memory)	0	0.0	0.0%	0.0	0.0%
P3: Asset Pipeline	61,719	9,241.7	25.6%	7,434.1	20.6%
P4: Bridge/Interop	0	0.0	0.0%	0.0	0.0%
Other	1,087	473.1	1.3%	455.8	1.3%

---

Top 20 Tasks by Self-Time

Self-time is the exclusive duration: time spent in the task itself, not in sub-tasks.
This is the most accurate indicator of where CPU cycles are actually spent.

Self (ms)	Inclusive (ms)	Count	Avg Self (us)	P95 Self (ms)	Max Self (ms)	% Work	Task Name	Top Caller
10,674.2	12,514.1	121,575	87.8	0.1	43.0	29.6%	module	write all entrypoints to disk (1%)
4,246.8	5,724.3	36,356	116.8	0.2	175.2	11.8%	analyze ecmascript module	process module (79%)
2,666.6	2,847.4	46,069	57.9	0.0	17.5	7.4%	internal resolving	resolving (29%)
2,491.1	6,120.6	57,059	43.7	0.0	17.4	6.9%	process module	module (10%)
2,088.1	2,984.7	37,315	56.0	0.0	25.9	5.8%	resolving	module (15%)
1,857.2	3,012.8	17,913	103.7	0.4	50.8	5.2%	code generation	chunking (2%)
1,708.9	1,968.0	7,154	238.9	0.7	58.3	4.7%	parse ecmascript	analyze ecmascript module (31%)
1,550.5	1,672.8	12,995	119.3	0.3	148.1	4.3%	chunking	write all entrypoints to disk (0%)
1,065.5	1,065.5	10,127	105.2	0.3	10.3	3.0%	precompute code generation	code generation (56%)
733.5	733.5	6,924	105.9	0.4	71.0	2.0%	compute async module info	chunking (0%)
657.9	914.5	7,632	86.2	0.0	118.8	1.8%	write all entrypoints to disk	None (0%)
484.3	484.8	9,797	49.4	0.0	35.1	1.3%	compute async chunks	webpack loader (0%)
412.7	423.4	979	421.5	1.8	22.1	1.1%	webpack loader	parse css (13%)
355.5	355.5	2,083	170.7	0.5	18.5	1.0%	generate source map	code generation (96%)
341.2	588.0	726	470.0	2.4	32.1	0.9%	parse css	module (7%)
271.2	271.2	2,338	116.0	0.2	15.1	0.8%	read file	parse ecmascript (91%)
49.2	49.2	818	60.1	0.0	8.4	0.1%	compute binding usage info	write all entrypoints to disk (0%)
27.9	27.9	1,810	15.4	0.0	6.2	0.1%	collect mergeable modules	compute merged modules (0%)
26.8	30.3	538	49.9	0.1	5.2	0.1%	async reference	write all entrypoints to disk (1%)
21.8	45.2	106	205.3	0.5	13.9	0.1%	make production chunks	chunking (4%)

---

Critical Path Analysis

The longest sequential dependency chains that determine wall-clock time.
Focus on reducing the depth of these chains to improve parallelism.

Rank	Self-Time (ms)	Depth	Path
1	175.2	2	process module → analyze ecmascript module
2	87.1	2	process module → analyze ecmascript module
3	69.3	2	code generation → generate source map
4	58.3	2	analyze ecmascript module → parse ecmascript
5	46.5	2	process module → analyze ecmascript module

---

Batching Candidates

High-volume tasks dominated by a single parent. If the parent can batch them,
it drastically reduces scheduler overhead.

Task Name	Count	Top Caller (Attribution)	Avg Self	P95 Self	Total Self
analyze ecmascript module	36,356	process module (79%)	116.8 us	0.16 ms	4,246.8 ms

---

Duration Distribution

Range	Count	Percentage
<10us	1,243,758	76.5%
10us-100us	355,642	21.9%
100us-1ms	18,890	1.2%
1ms-10ms	6,194	0.4%
10ms-100ms	434	0.0%
>100ms	5	0.0%

---

Action Items

1. [P0] Focus on tasks with the highest Self-Time — these are where CPU cycles are actually spent.
2. [P0] Use Batching Candidates to identify callers that should use try_join or reduce #[turbo_tasks::function] granularity.
3. [P1] Check Build Phase Timeline for phases with disproportionate wall range vs. self-time (= serialization).
4. [P1] Inspect P95 Self (ms) for heavy monolith tasks. Focus on long-tail outliers, not averages.
5. [P1] Review Critical Paths — reducing the longest chain depth directly improves wall-clock time.
6. [P2] If Thread Utilization < 60%, investigate scheduling gaps (lock contention or deep dependency chains).

Report generated by Utoopack Performance Analysis Agent