fix(minifier): cap if-return ternary collapse for firefox#21841
Merged
Boshen merged 7 commits intoMay 7, 2026
Merged
Conversation
Member
|
The problem sounds similar to swc-project/swc#11736, which I came across the other day. |
Contributor
Author
|
@Boshen gentle reminder for a review :) |
|
I tested Safari's limit to be 6325. Also, I think it would be good for the limit to be enabled by default. i.e. it's hard for me to imagine situations where people someone would explicitly choose to turn off the limit. |
…llapse
The peephole pass in minimize_statements.rs collapses a chain of
consecutive `if (cond) return x;` statements into a single right-leaning
nested ternary (`return cond1 ? x : cond2 ? y : ... : z;`). There is no
depth cap on that collapse, so a function with N such cases produces a
`ConditionalExpression` nested N-1 levels deep.
This is a problem for Firefox. SpiderMonkey's parser refuses to parse
right-leaning ternary chains above roughly 4000-5000 levels, throwing
`InternalError: too much recursion` at script compile time with no
filename or line info. Chromium's V8 parses 100000+ without trouble.
Real-world trigger: webpack's `__webpack_require__.u` with one `if` per
chunk, in a bundle with several thousand chunks.
This change adds a new `Option<u32>` option:
CompressOptions { max_conditional_depth: Option<u32>, ... }
When set, the `return_loop` and `throw_loop` each measure the depth of
the `ConditionalExpression` already held by the trailing return/throw
argument via `Self::conditional_alternate_depth`, and break out of the
merge loop once that depth reaches the cap. Remaining if-return and
if-throw statements stay as separate statements. Because the check
looks at the actual ternary depth (not a loop counter), it stays
correct across the minifier's re-entrant passes.
Default is `None`. Behaviour unchanged for existing callers.
Test-Plan:
- cargo test -p oxc_minifier, 489/489 pass including the new
test_max_conditional_depth_caps_return_ternary_chain covering the
uncapped baseline, cap=2 (return), cap=0, and cap=2 (throw).
- Option plumbed through napi/minify for the NAPI binding.
Merging this PR will not alter performance
Comparing Footnotes
|
d8d16d9 to
e4c1d7f
Compare
camc314
added a commit
that referenced
this pull request
May 11, 2026
### 🚀 Features - 66c9b01 transformer/typescript: Debug_assert that `enum_eval` ran in semantic (#22252) (Dunqing) - ffe6475 minifier: Fold `Array` constructor with safe spreads (#22215) (camc314) ### 🐛 Bug Fixes - d3d0b18 traverse: Handle `ChainElement::TSNonNullExpression` in `GatherNodeParts` (#22247) (leaysgur) - 4e880de transformer/object-rest-spread: Declare temp vars for computed keys (#22284) (camc314) - a7c3e22 semantic: Clear member write target for computed keys (#22302) (camc314) - 6a8852d codegen: Emit newline after legal-comment orphan flush (#22304) (Dunqing) - 5da9fda transformer/explicit-resource-management: Preserve class names (#22306) (Dunqing) - b5d970f transformer/explicit-resource-management: Preserve class names (#22290) (camc314) - bc54fd4 minifier: Keep function / class names if direct eval is present in the scope (#22241) (sapphi-red) - 7a810c0 minifier: Refresh direct eval flags after DCE (#21787) (Dunqing) - dd88726 transformer/legacy-decorator: Preserve accessor type annotation for emitDecoratorMetadata (#21966) (Dunqing) - 29a3cd7 codegen: Swap mapping/indent order for top-level decls (#22206) (Dunqing) - 73b4f40 minifier: Preserve catch binding with direct eval (#22221) (camc314) - 0e13d17 minifier: Preserve optional chain base side effects (#22219) (camc314) - 0c7c01c transformer/typescript: Inline optional-chain enum member access (#21834) (Dunqing) - a6aff7e codegen: Emit block/array/object end mapping at close char (#22200) (Dunqing) - a099b03 codegen: Emit call end mapping at `)` position, not past it (#22199) (Dunqing) - 5753774 minifier: Cap if-return ternary collapse for firefox (#21841) (Gurupungav Narayanan) - 2493bdd codegen: Correct sourcemap end mappings for closing delimiters (#22001) (Mark Dalgleish) - 3b385e2 minifier: Bail optimizing `Array` with unknown arg count (#22188) (camc314) - 9fa2122 parser: Parse array computed class keys (#22159) (camc314) ### 📚 Documentation - a4a6892 napi/parser: Correct code comment (#22278) (overlookmotel) - 9305373 oxc: Update README (#22178) (camc314) Co-authored-by: Cameron <[email protected]>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Firefox's SpiderMonkey parser refuses to parse a right-leaning ?: chain above roughly 4000-5000 levels and throws InternalError: too much recursion at script-compile time, with no filename or line number. V8 parses 100 000+ without trouble. This is reproducible standalone in a tiny repro repo:
https://github.com/guru-irl/firefox-ternary-parse-recursion (42 Playwright assertions across Chromium 147 and Firefox 148, three literal types, eight depths)
Mozilla bug: https://bugzilla.mozilla.org/show_bug.cgi?id=2034840