Reproduction link or steps
https://github.com/naruaway-sandbox/20260615-rolldown-nondeterministic-build-repro
What is expected?
When the build input is fixed, the build output should be fixed (deterministic)
What is actually happening?
Rolldown randomly choose a symbol name to be used in the build output in this case, which makes the build non deterministic
System Info
System:
OS: macOS 26.4.1
CPU: (16) arm64 Apple M4 Max
Memory: 15.47 GB / 64.00 GB
Shell: 5.9 - /bin/zsh
Binaries:
Node: 26.0.0 - /opt/homebrew/bin/node
Yarn: 1.22.22 - /opt/homebrew/bin/yarn
npm: 11.12.1 - /opt/homebrew/bin/npm
pnpm: 11.5.2 - /opt/homebrew/bin/pnpm
bun: 1.3.14 - /opt/homebrew/bin/bun
Deno: 2.6.6 - /opt/homebrew/bin/deno
Watchman: 2026.01.12.00 - /opt/homebrew/bin/watchman
Browsers:
Safari: 26.4
Any additional comments?
I just put "System Info" as instructed but the repro https://github.com/naruaway-sandbox/20260615-rolldown-nondeterministic-build-repro uses Pnpm with a fixed Pnpm/Node.js version (latest) for better reproducibility of the issue
Reproduction link or steps
https://github.com/naruaway-sandbox/20260615-rolldown-nondeterministic-build-repro
What is expected?
When the build input is fixed, the build output should be fixed (deterministic)
What is actually happening?
Rolldown randomly choose a symbol name to be used in the build output in this case, which makes the build non deterministic
System Info
System: OS: macOS 26.4.1 CPU: (16) arm64 Apple M4 Max Memory: 15.47 GB / 64.00 GB Shell: 5.9 - /bin/zsh Binaries: Node: 26.0.0 - /opt/homebrew/bin/node Yarn: 1.22.22 - /opt/homebrew/bin/yarn npm: 11.12.1 - /opt/homebrew/bin/npm pnpm: 11.5.2 - /opt/homebrew/bin/pnpm bun: 1.3.14 - /opt/homebrew/bin/bun Deno: 2.6.6 - /opt/homebrew/bin/deno Watchman: 2026.01.12.00 - /opt/homebrew/bin/watchman Browsers: Safari: 26.4Any additional comments?
I just put "System Info" as instructed but the repro https://github.com/naruaway-sandbox/20260615-rolldown-nondeterministic-build-repro uses Pnpm with a fixed Pnpm/Node.js version (latest) for better reproducibility of the issue