refactor: project the retained export interface out of used_symbol_refs by hyfdev · Pull Request #10089 · rolldown/rolldown

hyfdev · 2026-07-02T11:15:22Z

Description

Part 3 of the used_symbol_refs decomposition stack (stacked on #10088).

Consumers that iterate a module's resolved exports were answering "is this export retained" by querying used_symbol_refs directly. This PR projects that answer into a dedicated RetainedExportSymbols set — for every module's resolved_exports entry whose usage survived, the export's symbol_ref as recorded plus its canonical form — computed right after the module-namespace decision so export * as ns exports reflect it. Switched consumers:

rendered-module export lists (ecma_generator.rs)
the namespace object property list in the finalizer
entry-chunk export naming and AllowExtension emitted-chunk exports (compute_cross_chunk_links)

With both this and #10087 in place, the finalizer context's used_symbol_refs field had no readers left, so it is removed here.

Note: the ESM external-import fallback from #10088 deliberately does NOT move to this projection. An instrumented attempt changed 20 snapshots: the non-external canonical refs it sees include imports kept alive by eval and platform-import cases, which are no module's export, so the export projection cannot answer them.

No behavior change; snapshots are untouched.

Stack: #10087 → #10088 → #10089 → #10090 → #10091

hyfdev · 2026-07-02T11:15:36Z

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

This PR continues the used_symbol_refs decomposition by projecting “is this resolved export retained?” into a dedicated RetainedExportSymbols set computed during the generate stage, and migrating export-iteration consumers away from querying used_symbol_refs directly.

Changes:

Add RetainedExportSymbols (in rolldown_common) and store it on LinkStageOutput, computed after the module-namespace inclusion decision.
Switch export-interface consumers (module finalizer, rendered-module export list, cross-chunk export naming) to consult retained_export_symbols.
Remove the finalizer-context dependency on used_symbol_refs by threading retained_export_symbols instead.

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/mod.rs	Adds `retained_export_symbols` field to `LinkStageOutput` and initializes it.
crates/rolldown/src/stages/generate_stage/mod.rs	Computes retained export symbols immediately after namespace inclusion finalization.
crates/rolldown/src/stages/generate_stage/finalize_modules.rs	Passes `retained_export_symbols` into the finalizer context.
crates/rolldown/src/stages/generate_stage/compute_cross_chunk_links.rs	Uses `retained_export_symbols` when filtering/preserving exports for cross-chunk naming.
crates/rolldown/src/stages/generate_stage/code_splitting.rs	Implements `compute_retained_export_symbols` projection logic.
crates/rolldown/src/module_finalizers/mod.rs	Uses `retained_export_symbols` to decide whether to keep resolved exports (vs `used_symbol_refs`).
crates/rolldown/src/module_finalizers/finalizer_context.rs	Replaces finalizer-context `used_symbol_refs` with `retained_export_symbols`.
crates/rolldown/src/ecmascript/ecma_generator.rs	Filters rendered-module export list using `retained_export_symbols`.
crates/rolldown_common/src/types/retained_export_symbols.rs	Introduces the new retained-export set type.
crates/rolldown_common/src/types/mod.rs	Exposes the new `retained_export_symbols` module.
crates/rolldown_common/src/lib.rs	Re-exports `RetainedExportSymbols` from `rolldown_common`.

codspeed-hq · 2026-07-02T11:49:44Z

Merging this PR will not alter performance

✅ 7 untouched benchmarks
⏩ 10 skipped benchmarks¹

_{Comparing 07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs (7e29bd6) with main (fdb678e)²}

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. ↩
No successful run was found on graphite-base/10089 (31c8e52) during the generation of this report, so main (fdb678e) was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

hyfdev · 2026-07-02T13:00:17Z

Merge activity

Jul 2, 1:00 PM UTC: The merge label 'graphite: merge-when-ready' was detected. This PR will be added to the Graphite merge queue once it meets the requirements.
Jul 2, 1:02 PM UTC: hyf0 added this pull request to the Graphite merge queue.
Jul 2, 1:10 PM UTC: The Graphite merge queue couldn't merge this PR because it had merge conflicts.
Jul 2, 1:12 PM UTC: Graphite rebased this pull request, because this pull request is set to merge when ready.
Jul 2, 1:15 PM UTC: hyf0 added this pull request to the Graphite merge queue.
Jul 2, 1:21 PM UTC: Merged by the Graphite merge queue.

…a field (#10087) ### Description Part 1 of a 4-PR stack that gives "is this symbol used" a single authority per question, by splitting purpose-specific views out of `used_symbol_refs`. Whether a module's namespace object is retained was previously communicated to downstream passes by inserting/removing the module's `namespace_object_ref` in `used_symbol_refs` (#7002). That made the set a mutable side channel after tree-shaking had settled: from that point on the set and the statement-inclusion bits could disagree, and each reader picked one of them ad hoc. This PR records the decision on a dedicated field, `LinkingMetadata::namespace_included`, computed in `finalized_module_namespace_ref_usage` from `module_namespace_included_reason`, and switches the dedicated readers over: - the finalizer's namespace-declaration gate (`finalizer_context.rs`) - the JSON export-interface check in `compute_cross_chunk_links` The set insert/remove remains as a mirror derived from the flag, because generic symbol-usage queries can still receive namespace refs (e.g. a `resolved_export.symbol_ref` produced by `export * as ns from '...'`). No behavior change; snapshots are untouched. Stack: #10087 → #10088 → #10089 → #10090 → #10091

#10088) ### Description Part 2 of the `used_symbol_refs` decomposition stack (stacked on #10087). Symbols owned by external modules — an external module's `namespace_ref`, and the per-name facade symbols created by the external import binding merger in `bind_imports_and_exports` — have no statements, so their usage can only ever be tracked symbol-level. This PR gives them a dedicated `UsedExternalSymbols` set, written by `include_symbol` alongside `used_symbol_refs` whenever the inserted ref's owner is an external module, and switches the external-only consumers to it: - ESM import specifier rendering (`format/esm.rs`) - CJS/ESM external interop checks (`format/cjs.rs`, `format/utils/mod.rs`) - chunk-level deconflicting (`deconflict_chunk_symbols.rs`) The ESM named-import path keeps a `used_symbol_refs` fallback for canonical refs that stay importer-local: a named import that is itself re-exported skips the binding merger, so its canonical ref is not external-owned. (Part 3 initially tried to move that fallback to the export projection; it turned out to also cover eval-kept imports and platform-import cases that are no module's export, so it stays on the set — see the description there.) No behavior change; snapshots are untouched. Stack: #10087 → #10088 → #10089 → #10090 → #10091

netlify · 2026-07-02T13:11:51Z

✅ Deploy Preview for rolldown-rs canceled.

Name	Link
🔨 Latest commit	`4aa1c16`
🔍 Latest deploy log	https://app.netlify.com/projects/rolldown-rs/deploys/6a4664bbff663c000801e03f

…fs (#10089) ### Description Part 3 of the `used_symbol_refs` decomposition stack (stacked on #10088). Consumers that iterate a module's resolved exports were answering "is this export retained" by querying `used_symbol_refs` directly. This PR projects that answer into a dedicated `RetainedExportSymbols` set — for every module's `resolved_exports` entry whose usage survived, the export's `symbol_ref` as recorded plus its canonical form — computed right after the module-namespace decision so `export * as ns` exports reflect it. Switched consumers: - rendered-module export lists (`ecma_generator.rs`) - the namespace object property list in the finalizer - entry-chunk export naming and AllowExtension emitted-chunk exports (`compute_cross_chunk_links`) With both this and #10087 in place, the finalizer context's `used_symbol_refs` field had no readers left, so it is removed here. Note: the ESM external-import fallback from #10088 deliberately does NOT move to this projection. An instrumented attempt changed 20 snapshots: the non-external canonical refs it sees include imports kept alive by `eval` and platform-import cases, which are no module's export, so the export projection cannot answer them. No behavior change; snapshots are untouched. Stack: #10087 → #10088 → #10089 → #10090 → #10091

### Description Part 4 (last) of the `used_symbol_refs` decomposition stack (stacked on #10089). The original endgame was to delete `UsedSymbolRefs` entirely once every consumer had a purpose-specific home. An instrumented run over the whole test corpus rejected that: the cross-chunk filters (`compute_cross_chunk_links`) drop three populations of refs — - ~1.3k inlinable constants (absent from the set by design: their use sites hold the value) - ~0.7k namespace objects eliminated by the generate stage - **~1.3k over-collected depended refs that neither constants nor namespaces explain** — refs only the inclusion fixpoint's own record can rule dead The last population is not answerable by statement-inclusion bits (one bit per statement cannot distinguish which of a statement's bindings is referenced), nor by any of the new views. So the set stays — as the single record of fixpoint liveness. This PR writes that contract on the type: what membership means, the three sanctioned writers (the inclusion pass, the chunk optimizer's facade-elimination re-run, and the namespace-decision mirror), and pointers to the purpose-specific views (`LinkingMetadata::namespace_included`, `UsedExternalSymbols`, `RetainedExportSymbols`) that answer the common questions. The over-collected population also suggests `collect_depended_symbols` collects more than needed; that is left as a separate investigation. Docs/comments only; no behavior change. Stack: #10087 → #10088 → #10089 → #10090 → #10091

…#10091) ### Description Part 5 (follow-up to the review discussion) of the `used_symbol_refs` decomposition stack (stacked on #10090). #10090 documented by convention that `used_symbol_refs` must not change after the inclusion machinery finishes. This PR enforces it at the type level instead: - The mutable phase is a separate `UsedSymbolRefsBuilder` (`insert`/`contains`), held by the link-stage fixpoint and threaded explicitly through `generate_chunks` into the chunk optimizer's facade-elimination re-run — the last writer. `remove()` no longer exists at all. - `seal()` consumes the builder and yields `UsedSymbolRefs`, which only exposes `contains()`. Downstream passes (retained-export projection, cross-chunk linking, chunk output exports, chunk rendering) receive the sealed value. Mutating after the seal point is a compile error, not a runtime check. - The set no longer lives on `LinkStageOutput`; the ESM generator reads it through a dedicated `GenerateContext` field. This also deletes the namespace-decision mirror left by #10087: `finalized_module_namespace_ref_usage` now only writes `meta.namespace_included`, and the three places whose answers previously depended on the mirrored set membership for namespace refs — the two cross-chunk liveness filters and the retained-export projection — consult the flag explicitly. `UsedExternalSymbols` and `RetainedExportSymbols` could get the same builder/sealed split as a follow-up. No behavior change; snapshots are untouched. Stack: #10087 → #10088 → #10089 → #10090 → #10091

@V1OL3TF0X

## [1.1.5] - 2026-07-08 ### 🚀 Features - detect top-level import-binding reads as execution-order sensitive (#10180) by @hyf0 - sourcemap_filenames: add a sourcemapFileNames option (#9271) by @V1OL3TF0X - binding: record plugin hook result kind in tracing spans (#10154) by @IWANABETHATGUY - linking: skip side-effect-free modules in per-entry reachability (#10111) by @IWANABETHATGUY - improve error message for unresolved virtual imports (#10156) by @sapphi-red - add descriptive metadata to plugin API (#10106) by @sapphi-red - add `--configLoader=native` option (#10118) by @sapphi-red ### 🐛 Bug Fixes - improve invalid annotation warnings (#10185) by @hyf0 - keep deduplicated asset filenames stable once they can be observed (#10191) by @shulaoda - sourcemap_filenames: use public option name in pattern errors (#10188) by @IWANABETHATGUY - sourcemap_filenames: hash prepared sourcemap content (#10178) by @hyf0 - tree-shake unused circular declarators exported via export list (#10166) by @IWANABETHATGUY - dev: don't panic when an HMR rebuild hits an unresolved import (#10162) by @shulaoda - propagate errors from output.globals function (#9880) by @shulaoda - dev: revert cache mutations when a partial scan fails (#10110) by @shulaoda - dev: update importer relationships of cached modules in incremental build (#10107) by @shulaoda - hmr: fall back to full reload when a changed module is not registered as executed (#10132) by @shulaoda - chunk-optimizer: follow entry facade edges in runtime placement cycle check (#10101) by @hyf0 - dev: ignore watcher events after close (#10113) by @hyf0 - emit async wrapper for TLA modules under onDemandWrapping (#10086) by @IWANABETHATGUY - gate sideEffects:false modules' side effects on body demand (#10080) by @IWANABETHATGUY - rolldown_plugin_vite_resolve: return empty object for `browser: false` mapped modules (#10082) by @sapphi-red - reset the word-boundary state on newline in Hires::Boundary sourcemaps (#10025) by @shulaoda - trim an emptied chunk's outro/intro instead of skipping past it (#10029) by @shulaoda - test each edited chunk's own start against indent exclude ranges (#10026) by @shulaoda - preserve sourcemap mappings for indented lines when a CJS module shares the chunk (#10074) by @hyf0 ### 🚜 Refactor - separate tree-shaking side effects from execution order sensitivity (#10168) by @hyf0 - type construct_vite_preload_call to take an ObjectPattern (#10135) by @shulaoda - treeshake: single-source the own-export classification shared with the lazy-barrel loader (#10098) by @IWANABETHATGUY - dev: reuse Vite's bundledDev server (#10081) by @h-a-n-a - clippy: ban std HashMap/HashSet in favour of FxHashMap/FxHashSet (#10108) by @Boshen - treeshake: make body demand a second module bit instead of a stmt multimap (#10097) by @IWANABETHATGUY - seal used_symbol_refs by construction after its last writer (#10091) by @hyf0 - treeshake: replace inclusion mutual recursion with a worklist engine (#10096) by @IWANABETHATGUY - treeshake: split include_statements.rs into focused modules (#10095) by @IWANABETHATGUY - drop redundant is_user_defined filter on partitioned entries (#10050) by @shulaoda - project the retained export interface out of used_symbol_refs (#10089) by @hyf0 - track used external symbols separately from used_symbol_refs (#10088) by @hyf0 - make module namespace inclusion an explicit linking metadata field (#10087) by @hyf0 - rename statement evaluation metadata (#10078) by @hyf0 ### 📚 Documentation - virtual modules user-facing id convention (#10155) by @sapphi-red - cli: clarify disabling boolean/object flags like codeSplitting (#10153) by @IWANABETHATGUY - chore: remove Vite+ alpha banner (#10105) by @mdong1909 - write down the used_symbol_refs contract (#10090) by @hyf0 - dev/lazy: update design and implementation (#10079) by @h-a-n-a ### ⚡ Performance - ast_scanner: stop order-sensitivity checks once a module is flagged (#10190) by @IWANABETHATGUY - return impl ExactSizeIterator from slice-backed accessors (#10133) by @Boshen - binding: box dev and watcher napi futures (#10103) by @Boshen ### 🧪 Testing - move string_wizard replace unit tests to the JS magic-string suite (#10176) by @IWANABETHATGUY - dev: assert incremental scan state matches a fresh full build after each HMR step (#10115) by @shulaoda - dev: restore runtime assertions of delete_file_not_used_anymore (#10112) by @shulaoda - dev: fix flaky dev server tests in CI (#10152) by @h-a-n-a - add regression test for #10099 (lazyBarrel drops default-import binding but keeps its property reads) (#10109) by @IWANABETHATGUY ### ⚙️ Miscellaneous Tasks - deploy website to Void via GitHub OIDC (#10192) by @Boshen - deps: update oxc to 0.139.0 (#10161) by @shulaoda - deps: update test262 submodule for tests (#10160) by @rolldown-guard[bot] - rolldown_plugin_utils: remove dead asset-url and css scaffolding (#10131) by @shulaoda - deps: revert vite-plus to v0.2.1 (#10148) by @shulaoda - deps: update github actions (#10141) by @renovate[bot] - deps: update dependency rust to v1.96.1 (#10145) by @renovate[bot] - deps: update npm packages (#10142) by @renovate[bot] - deps: update rust crates (#10143) by @renovate[bot] - deps: update napi to v3.10.3 (#10121) by @renovate[bot] - rolldown_utils: remove unused time module (#10138) by @shulaoda - remove dead CopyModulePlugin::is_active method (#10129) by @shulaoda - remove dead LazyCompilationContext::is_lazy_module method (#10128) by @shulaoda - remove dead BuildDiagnostic::downcast_ref method (#10127) by @shulaoda - deps: update dependency vite-plus to v0.2.2 (#10084) by @renovate[bot] - deps: update rust crate oxc_sourcemap to v8.1.0 (#10122) by @renovate[bot] - deps: update crate-ci/typos action to v1.48.0 (#10124) by @renovate[bot] - enable more clippy restriction lints (#10114) by @Boshen - deps: update rust dependencies (#10100) by @Boshen - deps: update oxc resolver to v11.23.0 (#10083) by @renovate[bot] ### ◀️ Revert - Revert "chore(deps): revert vite-plus to v0.2.1" (#10157) by @h-a-n-a - "fix(hmr): fall back to full reload when a changed module is not registered as executed (#10132)" (#10151) by @shulaoda ### ❤️ New Contributors * @V1OL3TF0X made their first contribution in [#9271](#9271) Co-authored-by: shulaoda <[email protected]>

hyfdev mentioned this pull request Jul 2, 2026

refactor: make module namespace inclusion an explicit linking metadata field #10087

Merged

This was referenced Jul 2, 2026

refactor: track used external symbols separately from used_symbol_refs #10088

Merged

docs: write down the used_symbol_refs contract #10090

Merged

hyfdev force-pushed the 07-02-refactor_track_used_external_symbols_separately_from_used_symbol_refs branch from db6b082 to 89a8a14 Compare July 2, 2026 11:35

hyfdev force-pushed the 07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs branch from 97511cf to 45059f3 Compare July 2, 2026 11:35

hyfdev mentioned this pull request Jul 2, 2026

refactor: seal used_symbol_refs by construction after its last writer #10091

Merged

hyfdev marked this pull request as ready for review July 2, 2026 11:43

hyfdev requested review from IWANABETHATGUY and shulaoda as code owners July 2, 2026 11:43

Copilot AI review requested due to automatic review settings July 2, 2026 11:43

Copilot started reviewing on behalf of hyfdev July 2, 2026 11:43 View session

Copilot AI reviewed Jul 2, 2026

View reviewed changes

Comment thread crates/rolldown/src/stages/generate_stage/code_splitting.rs

hyfdev force-pushed the 07-02-refactor_track_used_external_symbols_separately_from_used_symbol_refs branch from 89a8a14 to 31c8e52 Compare July 2, 2026 11:56

hyfdev force-pushed the 07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs branch from 45059f3 to b65c6ee Compare July 2, 2026 11:56

IWANABETHATGUY approved these changes Jul 2, 2026

View reviewed changes

hyfdev force-pushed the 07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs branch from b65c6ee to 7e29bd6 Compare July 2, 2026 12:58

hyfdev added the graphite: merge-when-ready label Jul 2, 2026

graphite-app Bot changed the base branch from 07-02-refactor_track_used_external_symbols_separately_from_used_symbol_refs to graphite-base/10089 July 2, 2026 13:02

graphite-app Bot removed the graphite: merge-when-ready label Jul 2, 2026

graphite-app Bot changed the base branch from graphite-base/10089 to main July 2, 2026 13:08

graphite-app Bot added the graphite: merge-when-ready label Jul 2, 2026

hyfdev force-pushed the 07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs branch from 7e29bd6 to e36a655 Compare July 2, 2026 13:11

graphite-app Bot force-pushed the 07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs branch from e36a655 to 4aa1c16 Compare July 2, 2026 13:16

graphite-app Bot merged commit 4aa1c16 into main Jul 2, 2026
34 checks passed

graphite-app Bot removed the graphite: merge-when-ready label Jul 2, 2026

graphite-app Bot deleted the 07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs branch July 2, 2026 13:21

rolldown-guard Bot mentioned this pull request Jul 8, 2026

release: v1.1.5 #10194

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

refactor: project the retained export interface out of used_symbol_refs#10089

refactor: project the retained export interface out of used_symbol_refs#10089
graphite-app[bot] merged 1 commit into
mainfrom
07-02-refactor_project_the_retained_export_interface_out_of_used_symbol_refs

hyfdev commented Jul 2, 2026 •

edited

Loading

Uh oh!

hyfdev commented Jul 2, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

codspeed-hq Bot commented Jul 2, 2026 •

edited

Loading

Uh oh!

hyfdev commented Jul 2, 2026 •

edited by graphite-app Bot

Loading

Uh oh!

netlify Bot commented Jul 2, 2026 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

hyfdev commented Jul 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Uh oh!

hyfdev commented Jul 2, 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!

Uh oh!

codspeed-hq Bot commented Jul 2, 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!

hyfdev commented Jul 2, 2026 • edited by graphite-app Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merge activity

Uh oh!

netlify Bot commented Jul 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ Deploy Preview for rolldown-rs canceled.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

hyfdev commented Jul 2, 2026 •

edited

Loading

hyfdev commented Jul 2, 2026 •

edited

Loading

codspeed-hq Bot commented Jul 2, 2026 •

edited

Loading

hyfdev commented Jul 2, 2026 •

edited by graphite-app Bot

Loading

netlify Bot commented Jul 2, 2026 •

edited

Loading