feat: support inlineConst for CJS exports accessed through module.exports by h-a-n-a · Pull Request #8976 · rolldown/rolldown

h-a-n-a · 2026-04-01T03:59:13Z

Summary

When .default represents module.exports, properties accessed through it (e.g. ns.default.foo) can now be resolved and inlined by looking up the next property in the CJS module's exports.

Test plan

Extended esm_import_cjs_with_esmodule_flag fixture with ns.default.foo, ns.foo, named imports, and default imports
Added esm_import_cjs_with_esmodule_flag_non_const fixture for non-constant exports to verify the non-inlined path
All existing cjs_compat and inline_const tests pass

h-a-n-a · 2026-04-01T03:59:27Z

How to use the Graphite Merge Queue

Add the label graphite: merge-when-ready to this PR to add it to the merge queue.

You must have a Graphite account in order to use the merge queue. Sign up using this link.

_{An organization admin has enabled the Graphite Merge Queue in this repository.} _{Please do not merge from GitHub as this will restart CI on PRs being processed by the merge queue.}

This stack of pull requests is managed by Graphite. Learn more about stacking.

Copilot

Pull request overview

Adds CJS-interop support so that when .default represents module.exports (not exports.default), subsequent property accesses (e.g. ns.default.foo) can be resolved to the corresponding CJS export and eligible for inlineConst.

Changes:

Extend link-stage member-expression resolution to treat *.default.<prop> as a CJS export access when .default is module.exports.
Expand the existing esm_import_cjs_with_esmodule_flag fixture to cover ns.default.foo, ns.foo, and various import forms.
Add a new esm_import_cjs_with_esmodule_flag_non_const fixture to ensure non-constant CJS exports don’t get inlined.

Reviewed changes

Copilot reviewed 11 out of 11 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
crates/rolldown/src/stages/link_stage/bind_imports_and_exports.rs	Adds special-case resolution for `.default`-as-`module.exports` to enable resolving/inlining the next property as a CJS export.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag/main.js	Extends assertions to cover default/named imports and `.default.foo` access behavior.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag/cjs.cjs	Adds `exports.foo` to provide an additional CJS export for resolution/inlining.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag/cjs2.cjs	Adds a `module.exports = ...` case to test `.default` behavior when module.exports is non-object.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag/artifacts.snap	Updates snapshot to reflect new imports and inlining behavior.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag_non_const/_config.json	New fixture config enabling `optimization.inlineConst`.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag_non_const/package.json	New fixture set to Node ESM mode (`\"type\": \"module\"`).
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag_non_const/main.js	New assertions ensuring non-constant exports aren’t inlined.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag_non_const/cjs.cjs	New non-constant `exports.default` via `JSON.parse` to test non-inlined path.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag_non_const/cjs2.cjs	New non-constant `module.exports` via `JSON.parse` to test non-inlined path.
crates/rolldown/tests/rolldown/cjs_compat/esm_import_cjs_with_esmodule_flag_non_const/artifacts.snap	New snapshot for the non-const fixture.

Comments suppressed due to low confidence (1)

crates/rolldown/src/stages/link_stage/bind_imports_and_exports.rs:626

When resolving ns.default.<prop> where .default is treated as module.exports, the code still unconditionally depended_refs.push(m.symbol_ref) (and marks it written on assignment). Here m.symbol_ref is the CJS export for the current property (default), but that export is not actually what gets read/written in the module.exports case. This can force-include exports.default statements and/or incorrectly disable const inlining for default due to unrelated writes like ns.default.foo = .... Consider only adding m.symbol_ref to depended_refs/written_cjs_exports when that symbol is the actual resolution target; for the module.exports path, depend on the resolved next property (or namespace) instead.

                      target_commonjs_exported_symbol = Some((m.symbol_ref, is_default));
                    }
                    depended_refs.push(m.symbol_ref);
                    // If this member expression is a write (e.g. `cjs.c = 'abcd'`), the
                    // CJS exported symbol should not be inlined as a constant since its
                    // value may change at runtime.
                    if member_expr_ref.is_write {
                      written_cjs_exports.push(m.symbol_ref);
                    }

Copilot · 2026-04-01T04:06:13Z

crates/rolldown/src/stages/link_stage/bind_imports_and_exports.rs

+                      target_commonjs_exported_symbol = self.module_table[cjs_idx]
+                        .as_normal()
+                        .map(|m| (m.namespace_object_ref, true));


default_is_module_exports now sets target_commonjs_exported_symbol to the CJS module's namespace_object_ref when accessing just .default (no following property). Since include_statements skips the CJS bailout check whenever target_commonjs_exported_symbol is Some(_), this can incorrectly allow CJS tree-shaking even though .default represents the entire module.exports object (opaque use should bail out and keep all exports). Consider leaving target_commonjs_exported_symbol as None for the .default-as-module.exports case (or otherwise ensure bailout still happens for this specific scenario).

Suggested change

target_commonjs_exported_symbol = self.module_table[cjs_idx]

.as_normal()

.map(|m| (m.namespace_object_ref, true));

// `.default` here refers to the entire CJS `module.exports` object.

// Do NOT set `target_commonjs_exported_symbol` so that the CJS bailout

// logic still runs and keeps all exports for this opaque usage.

Nice catch. module.exports is an opaque access in this case. We have no idea which export it is accessing. So we should bail out in this case.

codspeed-hq · 2026-04-01T04:08:27Z

Merging this PR will not alter performance

✅ 4 untouched benchmarks
⏩ 10 skipped benchmarks¹

_{Comparing 04-01-feat_support_inlineconst_for_cjs_exports_accessed_through_module.exports (f1e6e0d) with main (76ce0cf)}

10 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

…orts

netlify · 2026-04-01T05:41:15Z

✅ Deploy Preview for rolldown-rs canceled.

Name	Link
🔨 Latest commit	`f1e6e0d`
🔍 Latest deploy log	https://app.netlify.com/projects/rolldown-rs/deploys/69ccafede93faf0008ea3273

github-actions bot assigned h-a-n-a Apr 1, 2026

h-a-n-a mentioned this pull request Apr 1, 2026

fix: correct inlining based on module's def format and esModule flag #8975

Merged

h-a-n-a marked this pull request as ready for review April 1, 2026 04:01

h-a-n-a requested review from IWANABETHATGUY and Copilot April 1, 2026 04:01

Copilot started reviewing on behalf of h-a-n-a April 1, 2026 04:01 View session

Copilot AI reviewed Apr 1, 2026

View reviewed changes

h-a-n-a force-pushed the 04-01-feat_support_inlineconst_for_cjs_exports_accessed_through_module.exports branch from b69cbfd to fb1f2f0 Compare April 1, 2026 04:13

h-a-n-a force-pushed the 04-01-fix_correct_inlining_based_on_module_s_def_format_and_esmodule_flag branch from 809c72c to 1c6712f Compare April 1, 2026 04:13

h-a-n-a force-pushed the 04-01-feat_support_inlineconst_for_cjs_exports_accessed_through_module.exports branch 2 times, most recently from 6bb9820 to d6eb912 Compare April 1, 2026 04:27

graphite-app bot changed the base branch from 04-01-fix_correct_inlining_based_on_module_s_def_format_and_esmodule_flag to graphite-base/8976 April 1, 2026 05:35

graphite-app bot force-pushed the graphite-base/8976 branch from 1c6712f to 76ce0cf Compare April 1, 2026 05:40

graphite-app bot force-pushed the 04-01-feat_support_inlineconst_for_cjs_exports_accessed_through_module.exports branch from d6eb912 to cb7e7ad Compare April 1, 2026 05:40

graphite-app bot changed the base branch from graphite-base/8976 to main April 1, 2026 05:40

feat: support inlineConst for CJS exports accessed through module.exp…

f1e6e0d

…orts

graphite-app bot force-pushed the 04-01-feat_support_inlineconst_for_cjs_exports_accessed_through_module.exports branch from cb7e7ad to f1e6e0d Compare April 1, 2026 05:40

h-a-n-a mentioned this pull request Apr 1, 2026

feat: treeshake module.exports property write #8659

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: support inlineConst for CJS exports accessed through module.exports#8976

feat: support inlineConst for CJS exports accessed through module.exports#8976
h-a-n-a wants to merge 1 commit intomainfrom
04-01-feat_support_inlineconst_for_cjs_exports_accessed_through_module.exports

h-a-n-a commented Apr 1, 2026 •

edited

Loading

Uh oh!

h-a-n-a commented Apr 1, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

h-a-n-a Apr 1, 2026 •

edited

Loading

Uh oh!

codspeed-hq bot commented Apr 1, 2026 •

edited

Loading

Uh oh!

netlify bot commented Apr 1, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

h-a-n-a commented Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Uh oh!

h-a-n-a commented Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

How to use the Graphite Merge Queue

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

h-a-n-a Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

codspeed-hq bot commented Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will not alter performance

Footnotes

Uh oh!

netlify bot commented Apr 1, 2026

✅ Deploy Preview for rolldown-rs canceled.

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

h-a-n-a commented Apr 1, 2026 •

edited

Loading

h-a-n-a commented Apr 1, 2026 •

edited

Loading

h-a-n-a Apr 1, 2026 •

edited

Loading

codspeed-hq bot commented Apr 1, 2026 •

edited

Loading