evmonly/evmonlyapp: revert the sized next-block decode (#4262), keep the PrepareBlock timer

[bdchatham]

#4262 sized PrepareBlock's decode of the next block from an EWMA of per-tx parse cost and half an EWMA of execution time, on the theory that decoding on every processor was contending with the current block's OCC speculation. The roll on testnet-2 (9a731d5) did not bear that out: steady-state executed tx/s was unchanged (93.9k vs 94.5k), and occ_speculate per tx barely moved once the estimate settled. Validators use ~24 of 32 cores, so speculation was never short of processors.

This reverts the sizing (PrepareBlockWithin, parseSizer, prepareBudget/nextExecuteEstimate, the executeEstimate on the app) so PrepareBlock decodes with the full ParseWorkers again, and re-adds the one part of #4262 that earned its keep: the evmonly_prepare phase timer around the decode in evmOnlyApplication.PrepareBlock, which is what shows prepare wall against block time on the dashboard.

No consensus or state impact; the decode result is identical either way. Existing evmonly and evmonlyapp tests pass with the race detector; the parse-sizer tests go with the code they covered.

Link to Devin session: https://app.devin.ai/sessions/ff612badcded4aa5914ea408dbb41888
Open in Devin Desktop: https://app.devin.ai/desktop/session/ff612badcded4aa5914ea408dbb41888?variant=devin
Requested by: @bdchatham

[devin-ai-integration]

I'll fix CI failures and address comments from users with write access. I'll skip comments containing "(aside)".

• Disable automatic comment, CI, and merge conflict monitoring

[cursor]

PR Summary

Low Risk
Rollback of parse-worker heuristics only; decode logic and consensus/state paths are unchanged aside from always using full ParseWorkers.

Overview
Reverts the budget-based, adaptive parse-worker sizing introduced in #4262 so next-block transaction decode in PrepareBlock again uses the full configured ParseWorkers count instead of a time-budgeted subset.

Removed: parseSizer and its tests, Executor.PrepareBlockWithin, and evmonlyapp helpers (prepareBudget, nextExecuteEstimate, executeEstimate) that sized decode from EWMA execution/parse cost. PrepareBlock is the single entry point again; prepared-path FinalizeBlock no longer records execution time for that sizing.

Kept: The evmonly_prepare phase timer around decode in evmOnlyApplication.PrepareBlock for dashboard visibility. Decode output is unchanged; this is a performance-tuning rollback with no consensus or state impact.

^{Reviewed by Cursor Bugbot for commit a2c0589. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

The latest Buf updates on your PR. Results from workflow Buf / buf (pull_request).

Build	Format	Lint	Breaking	Updated (UTC)
✅ passed	✅ passed	✅ passed	✅ passed	Sep 20, 2026, 3:21 PM

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.71%. Comparing base (7e3ca24) to head (a2c0589).
⚠️ Report is 45 commits behind head on giga-1.

Additional details and impacted files

@@             Coverage Diff             @@
##           giga-1    #4267       +/-   ##
===========================================
+ Coverage   65.55%   90.71%   +25.16%     
===========================================
  Files        2081       21     -2060     
  Lines      157460     3512   -153948     
===========================================
- Hits       103222     3186   -100036     
+ Misses      54097      326    -53771     
+ Partials      141        0      -141     

Flag	Coverage Δ
sei-chain	?
sei-chain-pr	90.71% <100.00%> (?)
sei-db	?

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
giga/evmonly/executor.go	91.28% <100.00%> (+0.07%)	⬆️
sei-tendermint/internal/evmonlyapp/app.go	86.86% <100.00%> (+5.09%)	⬆️

... and 2068 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[seidroid]

Clean, complete revert of the adaptive parse-worker sizing from #4262 that correctly retains the evmonly_prepare phase timer; no dangling references, no now-unused or now-missing imports, and the removed max(workers, 1) floor is still covered by Config.Sanitize plus parseBlockTxs's serial path. Only tests for the deleted code were removed.

Findings: 0 blocking | 0 non-blocking | 0 posted inline

Blockers

• None at the file/PR level.

Non-blocking

• None at the file/PR level.