nextjs - ✅(Solved) Fix Turbopack: dynamic cyclical import causes infinite loop [1 pull requests, 2 comments, 3 participants]

nextjs2025-10-20 17:56:35

ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

GitHub issue URL

Helpful · Quick feedback

GitHub stats

vercel/next.js#85119•Fetched 2026-04-08 02:17:22

View on GitHub

Comments

Participants

Timeline

Reactions

Author

Participants

Timeline (top)

referenced ×4labeled ×3subscribed ×3commented ×2

Root Cause

By selectively commenting out imports, I see that this roughly caused by the following import chain:

Fix Action

Fixed

Fixed by PR: Replace wasm_thread with web-thread (https://github.com/linera-io/linera-protocol/pull/4959)

PR fix notes

PR #4959: Replace `wasm_thread` with `web-thread`

Repository: linera-io/linera-protocol
Author: Twey
State: closed | merged: True
Link: https://github.com/linera-io/linera-protocol/pull/4959

Description (problem / solution / changelog)

Motivation

Our hacked-up version of wasm_thread isn't a great fit for our use cases.

Firstly, wasm_thread aims to mimic the API of std::thread. This was never sufficient for us as we need to send non-Send messages across threads actively using postMessage, which std::thread has no equivalent for. We patched wasm_thread but it was at odds with the point of the package.

Then, in order to use wasm_thread without threading a shim script URL through the code down into linera-base, we dynamically got the URL of the wasm-bindgen shim script using import.meta.url. This is poorly (and inconsistently) supported by bundlers, which want to know what the URL will be used for; but since the Worker construction in wasm_thread is done in Rust code they cannot analyze it. This resulted in our @linera/client library being rather fragile under different bundlers, with workarounds required to ensure they don't interfere with the paths of the various files.

Proposal

Introduce the web-thread library, a simplified equivalent to wasm_thread that has a Web-first API, including a trait for types that can be passed by postMessage (including transferable objects), and whose worker logic is primarily implemented in JavaScript, making it much more comprehensible to bundlers.

This allows us:

to communicate both Send and Post objects with long-running threads, as is required by https://github.com/linera-io/linera-protocol/issues/3651 and eventually https://github.com/linera-io/linera-protocol/issues/2927
to remove the bundling caveats from Linera documentation
to support WebPack
to delete linera_base::task and the conditional compilation therein

Test Plan

The web-thread library itself has been tested with Vite as well as WebPack (via Next.js with the --webpack flag), and includes example applications to that effect.

I have manually tested our two Web applications in this repository.

I defer to CI for full tests of the native pathways.

Release Plan

Nothing to do / These changes follow the usual release cycle.

Changed files

Cargo.lock (modified, +73/-50)
Cargo.toml (modified, +4/-5)
examples/Cargo.lock (modified, +60/-38)
examples/counter/package.json (modified, +1/-1)
examples/native-fungible/index.html (modified, +1/-1)
examples/native-fungible/package.json (modified, +1/-1)
examples/pnpm-lock.yaml (modified, +23/-13)
examples/vite.config.ts (modified, +0/-10)
linera-base/Cargo.toml (modified, +0/-7)
linera-base/src/lib.rs (modified, +1/-3)
linera-base/src/task.rs (modified, +13/-202)
linera-client/src/chain_listener.rs (modified, +5/-5)
linera-client/src/unit_tests/chain_listener.rs (modified, +2/-2)
linera-core/Cargo.toml (modified, +2/-0)
linera-core/src/chain_worker/actor.rs (modified, +45/-36)
linera-core/src/client/chain_client/mod.rs (modified, +1/-3)
linera-core/src/worker.rs (modified, +4/-4)
linera-execution/Cargo.toml (modified, +2/-1)
linera-execution/build.rs (modified, +1/-1)
linera-execution/src/execution.rs (modified, +6/-12)
linera-execution/src/execution_state_actor.rs (modified, +6/-12)
linera-execution/src/lib.rs (modified, +31/-29)
linera-execution/src/wasm/mod.rs (modified, +17/-22)
linera-sdk/tests/fixtures/Cargo.lock (modified, +31/-9)
linera-storage/Cargo.toml (modified, +1/-0)
linera-storage/src/lib.rs (modified, +10/-13)
web/@linera/client/package.json (modified, +1/-0)
web/@linera/client/tsconfig.json (modified, +2/-1)
web/pnpm-lock.yaml (modified, +3/-0)

Code Example

git clone https://github.com/maxpatiiuk/turbopack-dynamic-import-freeze.git
   cd turbopack-dynamic-import-freeze

---

npm install

---

npx next dev --turbopack
   # or:
   # CAUTION: the tracing file grows quickly, without limits
   NEXT_TURBOPACK_TRACING=1 npx next dev --turbopack

---

(also reproducible in Stackblitz)

Operating System:
  Platform: darwin
  Arch: arm64
  Version: Darwin Kernel Version 24.6.0: Mon Jul 14 11:30:29 PDT 2025; root:xnu-11417.140.69~1/RELEASE_ARM64_T6000
  Available memory (MB): 32768
  Available CPU cores: 10
Binaries:
  Node: 22.19.0
  npm: 10.9.3
  Yarn: 1.22.19
  pnpm: 8.9.0
Relevant Packages:
  next: 15.5.6 // Latest available version is detected (15.5.6).
  eslint-config-next: N/A
  react: 19.2.0
  react-dom: 19.2.0
  typescript: 5.9.3
Next.js Config:
  output: N/A

---

NEXT_TURBOPACK_TRACING=1 MallocNanoZone=0 NODE_ENV=development NEXT_RUNTIME=nodejs TURBOPACK=1 NEXT_PRIVATE_WORKER=1 NEXT_PRIVATE_TRACE_ID=cb4e155e9cf7b89b WATCHPACK_WATCHER_LIMIT=20 node --inspect-brk node_modules/next/dist/server/lib/start-server.js

RAW_BUFFERClick to expand / collapse

Link to the code that reproduces this issue

https://github.com/maxpatiiuk/turbopack-dynamic-import-freeze

To Reproduce

Clone the repository

git clone https://github.com/maxpatiiuk/turbopack-dynamic-import-freeze.git
cd turbopack-dynamic-import-freeze

OR: use the Stackblitz reproduction (but hard to debug in such environment):

Install dependencies
```
npm install
```

Run the development server

npx next dev --turbopack
# or:
# CAUTION: the tracing file grows quickly, without limits
NEXT_TURBOPACK_TRACING=1 npx next dev --turbopack

Current vs. Expected behavior

If a module dynamically imports another module, which in turn dynamically imports the first module, Turbopack gets into a loop (100% CPU usage, trace file grows indefinitely).

My best minimal reproduction is in app/page.tsx, where a single library entrypoint import causes the loop.

I also attached the trace-turbopack file: trace-turbopack.zip Note it grows indefinitely, so I killed the process shortly after starting it to make the file uploadable to GitHub.

Provide environment information

(also reproducible in Stackblitz)

Operating System:
  Platform: darwin
  Arch: arm64
  Version: Darwin Kernel Version 24.6.0: Mon Jul 14 11:30:29 PDT 2025; root:xnu-11417.140.69~1/RELEASE_ARM64_T6000
  Available memory (MB): 32768
  Available CPU cores: 10
Binaries:
  Node: 22.19.0
  npm: 10.9.3
  Yarn: 1.22.19
  pnpm: 8.9.0
Relevant Packages:
  next: 15.5.6 // Latest available version is detected (15.5.6).
  eslint-config-next: N/A
  react: 19.2.0
  react-dom: 19.2.0
  typescript: 5.9.3
Next.js Config:
  output: N/A

Which area(s) are affected? (Select all that apply)

Module Resolution

Which stage(s) are affected? (Select all that apply)

next dev (local)

Additional context

I tried starting the server inside a debugger. The best I can tell the .js side is idle while native code is stuck in a loop:

NEXT_TURBOPACK_TRACING=1 MallocNanoZone=0 NODE_ENV=development NEXT_RUNTIME=nodejs TURBOPACK=1 NEXT_PRIVATE_WORKER=1 NEXT_PRIVATE_TRACE_ID=cb4e155e9cf7b89b WATCHPACK_WATCHER_LIMIT=20 node --inspect-brk node_modules/next/dist/server/lib/start-server.js

By selectively commenting out imports, I see that this roughly caused by the following import chain:

app/page.tsx imports @arcgis/core/portal/Portal
Portal.js has import("../Basemap")
Basemap.js has import { getLayerJSON as _ } from './webdoc/support/writeUtils.js';
writeUtils.js has import { isFeatureCollectionLayer as i } from '../../layers/support/layerUtils.js';
layerUtils.js has import { isLayerFromCatalog as r } from '../catalog/catalogUtils.js';
catalogUtils.js has import n from '../Layer.js';
Layer.js has import { fromPortalItem as m } from './support/fromPortalItem.js'; and import('./support/arcgisLayers.js') - either causes the issue
fromPortalItem.js has import('../../portal/support/portalLayers.js')
portalLayers.js has import { findAssociatedFeatureService as a } from '../../layers/support/associatedFeatureServiceUtils.js';
associatedFeatureServiceUtils.js has import o from '../../portal/Portal.js';

However, reducing the chain to a smaller reproduction is tricky. I hope the provided trace file can help you debug the root cause.

extent analysis

TL;DR

The issue can be fixed by refactoring the import chain to avoid circular dependencies between modules.

Guidance

Identify and refactor circular imports: Analyze the provided import chain and identify the modules that are causing the circular dependency. Refactor the code to avoid these circular imports.
Use lazy loading or dynamic imports: Consider using lazy loading or dynamic imports to load modules only when needed, which can help avoid circular dependencies.
Verify the fix: Run the development server with the refactored code and monitor the CPU usage to ensure the issue is resolved.
Analyze the trace file: Use the provided trace file to gain insights into the import chain and identify potential areas for optimization.
Test with a smaller reproduction: Try to create a smaller reproduction of the issue to isolate the root cause and test the fix.

Example

No code example is provided as the issue is related to a complex import chain and requires a deeper analysis of the codebase.

Notes

The provided trace file and import chain suggest a complex issue that requires careful analysis and refactoring. The fix may involve significant changes to the codebase, and it's essential to test the changes thoroughly to ensure the issue is fully resolved.

Recommendation

Apply a workaround by refactoring the import chain to avoid circular dependencies. This approach is recommended as it addresses the root cause of the issue and provides a long-term solution.

Vote matrix · Quick signals

Works

Did the solution work? Tap to confirm.

Easy Fix

Was it a quick fix?

Time Saver

Did it save you time?

Blocking

Was it severely blocking?

Common Issue

Are others likely hitting this too?

Flaky / Intermittent

Is it intermittent?

Verified / Reproducible

Can you reproduce it reliably?

#prompt template #agent execution #callback error #memory management #API rate limit

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.