refactor(internals): reduce flushPending overhead (performance)

[dmaskasky]

Summary

• Adds an early return when there is no work to do
• Replaces forEach with for..of loops - slightly better performance?

Check List

• pnpm run fix for formatting and linting code and docs

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
jotai	Ready	Preview, Comment	Apr 3, 2026 6:09am

[codesandbox-ci]

This pull request is automatically built and testable in CodeSandbox.

To see build info of the built libraries, click here or the icon next to each commit SHA.

[dmaskasky]

FlushPending isn't all that hot. We can close this PR if we don't like the readability.

[pkg-pr-new]

More templates

• hacker_news
• hello
• mega-form
• starter
• text_length
• todos
• todos_with_atomFamily

npm i https://pkg.pr.new/jotai@3285

commit: c29f90f

[github-actions]

Preview in LiveCodes

Latest commit: c29f90f
Last updated: Apr 3, 2026 6:09am (UTC)

Playground	Link
React demo	https://livecodes.io?x=id/DARQFBG5R

---

See documentations for usage instructions.

[dai-shi]

I think changing do-while loop to while loop should be more readable and improve the performance for most cases. Only the concern is the store hook, right?

[dmaskasky]

Yeah. Also, btw flushCallbacks is not a hot path. Let me see if I can switch to a while.

[dmaskasky]

its ugly. I think I prefer the code as-written.

...
if (storeHooks.f) {
  call(storeHooks.f)
}
let iterations = 0
while (changedAtoms.size || unmountCallbacks.size || mountCallbacks.size) {
  if (storeHooks.f && iterations++ > 0) {
    call(storeHooks.f)
  }
  ...
}