fix(pack): skip ecmascript transforms for URL references

[fireairforce]

Summary

• exclude every URL reference subtype from ECMAScript transform rules
• add a generic dependency snapshot covering service worker and worker new URL(..., import.meta.url) usage from node_modules
• remove the stale invalid-module-type issue snapshots that the corrected rule no longer emits

Align the bahaviour with next.js turbopack: https://github.com/fireairforce/next.js/blob/canary/crates/next-core/src/next_shared/transforms/mod.rs#L108-L120

Validation

Update the snapshot test case

[gemini-code-assist]

Code Review

This pull request updates the JS module rule matching logic in crates/pack-core/src/shared/transforms/mod.rs to match any URL reference type by replacing Some(UrlReferenceSubType::Undefined) with None. Additionally, it removes obsolete error snapshot files and introduces a new snapshot test, foreign_js_new_url, to verify the handling of new URL references in external modules. I have no further feedback to provide as there are no review comments.

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.

[github-actions]

📊 Performance Benchmark Report (with-antd)

Utoopack Performance Report

Report ID: utoopack_performance_report_20260716_093656
Generated: 2026-07-16 09:36:56
Trace File: trace_antd.json (0.4GB, 1.13M spans)
Test Project: examples/with-antd

---

Executive Summary

Metric	Value	Assessment
Total Wall Time	7,679.4 ms	Baseline
Total Thread Work (de-duped)	22,036.0 ms	Non-overlapping busy time
Effective Parallelism	2.9x	thread_work / wall_time
Working Threads	9	Threads with actual spans
Thread Utilization	31.9%	⚠️ Suboptimal
Total Spans	1,128,144	All B/E + X events
Meaningful Spans (>= 10us)	367,922	(32.6% of total)
Tracing Noise (< 10us)	760,222	(67.4% 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	95,919	2,758.6	2,146.0	3,838.8
Parse	10,119	1,238.9	1,167.2	7,139.9
Analyze	234,876	13,289.5	9,356.9	7,022.7
Chunk	8,701	3,566.2	1,098.4	2,848.2
Codegen	12,072	1,745.3	1,664.6	2,450.2
Emit	26	39.2	19.4	7.8
Other	6,209	6,497.7	3,745.9	7,679.4

---

Workload Distribution by Diagnostic Tier

Category	Spans	Inclusive (ms)	% Work	Self-Time (ms)	% Self
P0: Scheduling & Resolution	333,396	16,463.2	74.7%	11,767.7	53.4%
P1: I/O & Heavy Tasks	2,903	121.5	0.6%	101.8	0.5%
P2: Architecture (Locks/Memory)	0	0.0	0.0%	0.0	0.0%
P3: Asset Pipeline	29,439	6,596.6	29.9%	3,976.5	18.0%
P4: Bridge/Interop	0	0.0	0.0%	0.0	0.0%
Other	2,184	5,953.8	27.0%	3,352.4	15.2%

---

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
4,919.9	7,036.1	148,659	33.1	0.1	13.1	22.3%	module	write all entrypoints to disk (1%)
2,090.0	2,129.0	26,006	80.4	0.2	258.7	9.5%	analyze ecmascript module	module (65%)
2,059.9	2,075.3	25	82395.1	332.2	503.0	9.3%	save snapshot	persist (4%)
1,366.3	1,497.3	53,101	25.7	0.0	8.9	6.2%	internal resolving	resolving (29%)
1,286.9	3,035.7	50,957	25.3	0.0	6.0	5.8%	process module	module (21%)
1,095.2	1,166.9	7,940	137.9	0.5	74.8	5.0%	parse ecmascript	process module (25%)
940.1	940.1	9,506	98.9	0.4	7.1	4.3%	precompute code generation	generate merged code (44%)
911.3	3,377.6	6,848	133.1	0.1	62.3	4.1%	chunking	compute async chunks (0%)
877.0	877.0	6,555	133.8	0.5	146.2	4.0%	compute async module info	compute merged modules (0%)
823.1	2,432.0	1,879	438.0	1.1	288.9	3.7%	generate merged code	chunking (52%)
769.4	1,250.9	42,120	18.3	0.0	4.2	3.5%	resolving	module (32%)
503.2	503.2	329	1529.6	1.5	299.1	2.3%	generate source map	code generation (83%)
429.0	883.7	175	2451.2	4.2	237.7	1.9%	emit code	generate merged code (33%)
359.7	507.3	3,593	100.1	0.0	198.4	1.6%	write all entrypoints to disk	None (0%)
221.3	301.9	2,237	98.9	0.2	50.1	1.0%	code generation	chunking (12%)
178.4	179.8	1,748	102.0	0.2	35.5	0.8%	compute async chunks	write all entrypoints to disk (0%)
118.3	118.3	726	162.9	0.0	40.3	0.5%	compute binding usage info	write all entrypoints to disk (1%)
71.5	71.5	2,168	33.0	0.1	2.7	0.3%	read file	parse ecmascript (91%)
51.9	51.9	1,913	27.1	0.0	11.9	0.2%	collect mergeable modules	compute merged modules (2%)
33.7	36.5	432	78.1	0.1	5.0	0.2%	async reference	write all entrypoints to disk (1%)

---

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	526.6	3	chunking → generate merged code → emit code
2	511.2	3	persist → save snapshot → blocking
3	366.5	4	chunking → generate merged code → emit code → generate source map
4	258.8	2	process module → analyze ecmascript module
5	143.5	2	save snapshot → blocking

---

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
No obvious batching candidates found	-	-	-	-	-

---

Duration Distribution

Range	Count	Percentage
<10us	760,222	67.4%
10us-100us	347,097	30.8%
100us-1ms	17,187	1.5%
1ms-10ms	3,510	0.3%
10ms-100ms	109	0.0%
>100ms	19	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