#36254 bench: add and use undo-data for BlockToJsonVerbosity3 bench

ReviewabilityReviewsAgreementSize
#36254 0xB10C bench: add and use undo-data for BlockToJsonVerbosity3 benchReady1Strong0 + 21 tests
  • Make the blockToJSON verbosity 3 benchmark accurately measure performance by including block undo data
  • Prevents verbosity 3 benchmarks from performing identical work to verbosity 2 without exercising prevouts
  • Ready to review
  • Strong support with no open objections
  • Verified by reproducing benchmark results and checking the undo data blob (ismaelsadeeq)
  • 0 lines added or modified outside tests
  • 21 lines added or modified in tests
  • 0 lines removed in total
  • 3 files, 1 commits
Test infrastructure#24 of 54
P3 · test coverage
  • P3 because it fixes an inaccurate benchmark harness to reflect true runtime performance
  • Ensures the benchmark actually exercises undo data reading, fee calculation, and prevout serialization
Ranking notes
Reviewability: ReadyStalePaused
Agreement: StrongPositivePositive w/ caveatsNeutralMildDisputedBlockedCrickets
Reviews: current code-review ACKs, then (+stale ACKs) and -NACKs; greener = more ACKs.
Size: lines added or modified outside tests, then in tests; greener = smaller.

Goal

Add undo data for block 413567 to benchmark test data and write it during test block index setup. This ensures that the RPC blockchain benchmarks for blockToJSON at verbosity level 3 exercise undo data reading, fee calculation, and prevout serialization.

Problem: The `blockToJSON` RPC helper requires undo data to populate prevouts and calculate transaction fees. Because the benchmark harness lacked undo data for the sample block, `TxVerbosity::SHOW_DETAILS` (verbosity 2) and `TxVerbosity::SHOW_DETAILS_AND_PREVOUT` (verbosity 3) performed identical work, giving an inaccurate benchmark of verbosity 3 performance.

Evidence of importance:

  • PR description benchmarks show that before the change, `BlockToJsonVerbosity2` ran in 17.07 ms/op and `BlockToJsonVerbosity3` ran in 17.25 ms/op (virtually identical), whereas with undo data available verbosity 3 increases to 34.27 ms/op as it processes prevouts.

Scope: Adds a binary raw data file (`src/bench/data/block413567_undo.raw`, ~136 KB).

Reviewability

Ready

Nothing known that would invalidate a review now.

Agreement

Strong

substantive support, no open objection (ismaelsadeeq)

Thread notes: Initial record. ismaelsadeeq provided a Tested ACK after reproducing benchmark results and verifying the undo-data blob; sedited offered an off-topic/related PR reference.

Objections, suggestions, and questions: none recorded.

Support:

  • ismaelsadeeq (member) (Tested ACK, 2026-09-15): I reproduced the benchmark and verified the undo-data blob.

Participants:

  • ismaelsadeeq (member, support, 1 statements 2026-09-15 to 2026-09-15): Tested ACK; reproduced benchmarks and verified undo-data blob
  • sedited (member, neutral, 1 statements 2026-09-15 to 2026-09-15): pointed to related PR 32554

Files

21 lines under test/bench/ci.

  • src/bench/rpc_blockchain.cpp +20/-0
  • src/bench/CMakeLists.txt +1/-0
  • src/bench/data/block413567_undo.raw +None/-None

Card

The short text the ranking pass compares across PRs.

This PR adds undo data for block 413567 to benchmark fixtures so that `blockToJSON` can read prevouts and calculate fees during benchmarks. Previously, the lack of undo data meant `BlockToJsonVerbosity3` (requesting prevouts) performed the exact same work as verbosity 2, masking the true cost of prevout extraction and serialization. Benchmark data in the PR description shows verbosity 3 runtime jumping from ~17.2 ms/op (identical to verbosity 2) to ~34.3 ms/op once undo data is actually processed.

Data

dossier JSON · extract JSON · model ledger, generated 2026-09-18T03:48, confidence high, input hash 3b9dee682dcf80b9
thread read through 2026-09-15 (3 statements), head 0771dd89d6
Discussion read under schema 1 rules: suggestions and questions may be missing, and an objection may combine several harms (schema history).
labels: Tests

Categories

Test infrastructure

P3 · test coverage

Fixes an inaccurate benchmark so that verbosity 3 actually exercises undo-data reading and prevout serialization instead of duplicating verbosity 2 work. The benchmark previously ran in 17.25 ms/op vs 17.07 ms/op for verbosity 2, whereas providing the undo data reflects its true ~34.27 ms/op runtime.

Membership: Touches src/bench/ and carries the Tests label, modifying the RPC blockchain benchmark harness and test data.

Factors: security/stability 0, bug 0, performance 0, user value 0, leverage 1