Rolldown v1.2.8 is out đ
Custom chunking with repeated `getModuleInfo` calls was 4.9à faster with 34% lower peak memory in our benchmark âĄ
- Speed up repeated `getModuleInfo` calls during custom chunking
- Fix Vite bundled-dev server crash after removing a dynamic import
- Fix external wildcard re-exports missing from `OutputChunk.imports`