feat: merge-train/fairies-v5 - #25329
Queued
AztecBot wants to merge 4 commits into
Queued
Conversation
## Summary The TXE bundle size guard (`enforceSizeLimits` in `yarn-project/txe/esbuild/plugins/size_guard.mjs`) summed **every** esbuild `metafile.outputs` entry — including the external `.map` sourcemaps — against `totalLimitMiB`. Since `esbuild.config.mjs` sets `sourcemap: 'external'` specifically so the runtime bundles stay small, counting `.map` files meant the guard mostly measured sourcemap size, not the runtime bundle. Measured on the current bundle: | | Size | | -- | -- | | Runtime JS (`.js`) | 4.74 MiB | | Sourcemaps (`.map`) | 9.32 MiB | | Old counted total | 14.06 MiB | So ~66% of the guarded number was sourcemaps. This is what tipped the guard over its limit on a sub-KB per-contract bytecode change. ## Changes - `enforceSizeLimits` now tracks two totals: runtime JS (`.js` outputs) and sourcemaps (`.map` outputs). The JS total is checked against `totalLimitMiB`; the sourcemap total against a new, deliberately loose `sourcemapLimitMiB`. - Reset `totalLimitMiB` `15 → 6` MiB (~1.3 MiB / ~27% headroom over the ~4.74 MiB of JS today). Reverts the 14.5 stopgap. - Added `sourcemapLimitMiB = 12` MiB (~2.7 MiB headroom over ~9.3 MiB today) so map growth is still watched, per the issue's optional suggestion. - Bump log gets a 2026-07-20 entry documenting the semantic switch to JS-only; header docstring rewritten to explain the two-total scheme. - The build's info line now prints `X MiB JS + Y MiB sourcemaps` instead of one combined figure. ## Verification Guard logic exercised against metafiles mirroring real sizes: - Current bundle → pass - +4 KB per-contract change (the scenario that tripped the guard) → pass - Sourcemaps grown past 12 MiB → trips *sourcemap total* only - Real JS grown past 6 MiB → trips *runtime JS total* - Sourcemaps never counted toward the JS total A real end-to-end `node ./esbuild.config.mjs` run also passed without tripping the guard. Closes F-812.
## Summary `BatchCall.simulate()` silently discarded `options.fee.paymentMethod` and dropped the merged payload's `feePayer`, so batches simulated with a sponsored/deferred fee method ran with the wallet's fallback handling (`PREEXISTING_FEE_JUICE`) instead of the caller's method. The wallet's account entrypoint then self-elected as fee payer and ended setup up front — a different phase structure than the caller asked for. Resolves F-785. Surfaced while landing #24479. ## Changes - `batch_call.ts`: fetch `options.fee.paymentMethod.getExecutionPayload()`, merge it ahead of the batch (mirroring `request()`), and preserve `combinedPayload.feePayer` when rebuilding the `ExecutionPayload`. - The wallet-side app-call offset only counts the wallet's own fee payment method, not one supplied via `options.fee`. So each batched call's private/public return-value index is offset by the fee payload's own call counts of each type; otherwise a call-contributing method (e.g. `SponsoredFeePaymentMethod`) shifts the indices and the fee call's return values decode as the first app call's. - `contract_function_interaction.ts`: fixed the same latent off-by-one under a call-contributing fee method. The single app call is the last of its type in the payload, so its index is derived from the merged payload (avoiding a second `getExecutionPayload()` call). - `phase_check.parallel.test.ts`: reverted the #24479 workaround back to the plain `batch.simulate({ from, fee })` form now that the option is honored. ## Testing - Added 3 `batch_call` unit tests: return-value index offset (asserting on actual decoded values, not just absence of errors), `feePayer` from the payment method, and `feePayer` carried by a batched payload. Verified red/green — the offset test fails without the fix. - All 3 `phase_check` e2e tests pass, including the reverted late-election test. - No behavior change for batches simulated without `fee.paymentMethod` (offsets are 0).
nventuro
added this pull request to the merge queue
Sep 1, 2026
Any commits made after this event will not be merged.
Collaborator
Author
|
🤖 Auto-merge enabled after 4 hours of inactivity. This PR will be merged automatically once all checks pass. |
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
4 participants
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.
See merge-train-readme.md.
This is a merge-train.