Skip to content

Embed created contracts as their standalone bytecode - #1632

Merged
sbillig merged 1 commit into
masterfrom
fix/embedded-contract-bytecode-consistency
Sep 29, 2026
Merged

sbillig merged 1 commit into
masterfrom
fix/embedded-contract-bytecode-consistency

Conversation

@cburgdorf

@cburgdorf cburgdorf commented Sep 29, 2026 •

Copy link
Copy Markdown
Collaborator

The init code a contract deploys with create<B>/create2<B> differed from the B.bin artifact written by fe build. All objects of a runtime package were lowered into one Sonatina module, and when B was embedded by a contract in another module it was recompiled inside that contract's module. Module-wide state then changed B's code: symbol conflict resolution renamed B's functions (changing the call-graph emission order), shared helpers got extra callers (changing inliner call counts and dead-argument elimination), and the inliner's module growth budget was shared. CREATE2 addresses derived from keccak(B.bin) therefore did not match the deployed address.

Bytecode emission now compiles every object in its own Sonatina module. Objects that a section embeds from another object are compiled first and placed into the embedding section as constant data; code region offset/length builtins for them resolve to that data. The embedded bytes are thus exactly the embedded object's own artifact, and an object's code no longer depends on other contracts in the same package. Test bytecode and trace emission use the same isolated-object path. Each traced object keeps its own pre/post-optimization identities, and regression tests compare trace hashes and test-embedded init code with standalone build artifacts.

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Sep 29, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-09-29T15:33:42.741229Z edd343a New commits
🔒 Security Review ✅ Completed 2026-09-29T15:30:46.055054Z edd343a New commits
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: c1970481ab

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread crates/codegen/src/sonatina/mod.rs Resolved
Comment thread crates/codegen/src/sonatina/mod.rs Resolved
The init code a contract deploys with `create<B>`/`create2<B>` differed
from the `B.bin` artifact written by `fe build`. All objects of a runtime
package were lowered into one Sonatina module, and when B was embedded by
a contract in another module it was recompiled inside that contract's
module. Module-wide state then changed B's code: symbol conflict
resolution renamed B's functions (changing the call-graph emission
order), shared helpers got extra callers (changing inliner call counts
and dead-argument elimination), and the inliner's module growth budget
was shared. CREATE2 addresses derived from keccak(B.bin) therefore did
not match the deployed address.

Bytecode emission now compiles every object in its own Sonatina module.
Objects that a section embeds from another object are compiled first and
placed into the embedding section as constant data; code region
offset/length builtins for them resolve to that data. The embedded bytes
are thus exactly the embedded object's own artifact, and an object's code
no longer depends on other contracts in the same package. The trace path
keeps compiling one module as before.
@cburgdorf
cburgdorf force-pushed the fix/embedded-contract-bytecode-consistency branch from c197048 to edd343a Compare September 29, 2026 15:27
@cburgdorf
cburgdorf requested a review from sbillig September 29, 2026 16:40
@sbillig
sbillig merged commit 7fdceda into master Sep 29, 2026
9 checks passed
@sbillig
sbillig deleted the fix/embedded-contract-bytecode-consistency branch September 29, 2026 17:34
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants