feat(spec-specs, tests): EIP-8037 - CREATE failure refunds state gas to reservoir

[spencer-tb]

🗒️ Description

Implements CREATE/CREATE2 failure refunds the account state gas charge (GAS_NEW_ACCOUNT) to state_gas_reservoir (STATE_BYTES_PER_NEW_ACCOUNT * cost_per_state_byte).

EIPs change for reference: ethereum/EIPs#11532

Spec change: 2be49b2

CREATE/CREATE2 keeps pay before execute (GAS_NEW_ACCOUNT charged upfront). On any failure path, the charge is refunded to state_gas_reservoir (and execution_state_gas_used decremented). Covers:

Silent failures (insufficient balance, nonce overflow, stack depth), address collision, child frame revert, child exceptional halt (including code deposit OOG and EIP 3541 invalid prefix).

Tests: 136312e

Six new tests in test_state_gas_create.py covering every failure branch:

• test_create_silent_failure_refunds_state_gas, parametrized failure_mode=[nonce_overflow, insufficient_balance]. Silent failures refund GAS_NEW_ACCOUNT, verified via header_verify on block gas_used.
  -test_create_child_revert_refunds_state_gas, parametrized gas_limit_mode=[with_reservoir, spillover] × create_opcode=[CREATE, CREATE2]. Child REVERT refunds the parent's CREATE charge. The spillover variant runs with tx gas at the cap so the state charge spills into gas_left, then the refund returns to the reservoir (not back to gas_left). Verified via header_verify.
• test_create_child_halt_refunds_state_gas, parametrized failure_mode=[initcode_halt, invalid_prefix] × create_opcode=[CREATE, CREATE2]. Exceptional halts consume all forwarded gas as regular_gas_used so block accounting cannot strictly discriminate via header gas. Uses a caller wrapper plus tight gas tuning pattern so the probe SSTORE can only succeed via the refunded reservoir.
• test_create_mixed_success_and_failure_block_accounting, parametrized create_opcode=[CREATE, CREATE2]. Successful CREATE followed by failed CREATE; block state gas reflects only the successful charges via header_verify.
• test_create_collision_refunds_state_gas, parametrized create_opcode=[CREATE, CREATE2] with @pre_alloc_mutable. A contract pre deployed at the CREATE/CREATE2 target triggers the account_has_code_or_nonce silent failure branch. Tight gas tuning via a caller wrapper leaves the factory in the discrimination window so the probe SSTORE succeeds only via the refund.
• test_create_code_deposit_oog_refunds_state_gas, parametrized create_opcode=[CREATE, CREATE2]. Initcode returns MAX_CODE_SIZE + 1 bytes, triggering an exceptional halt in code deposit. Tight gas tuning leaves the factory in the discrimination window so the probe SSTORE succeeds only via the refund.

🔗 Related Issues or PRs

• EIP text change: Update EIP-8037: more state gas accounting changes EIPs#11532

✅ Checklist

• All: Ran fast static checks to avoid unnecessary CI fails, see also Code Standards and Enabling Pre-commit Checks:

  just static

• All: PR title adheres to the repo standard and will be used as the squash commit message, starting with type(scope):.
• All: Considered updating the online docs in the ./docs/ directory.
• All: Set appropriate labels for the changes (only maintainers can apply labels).

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (eips/amsterdam/eip-8037@cef0fde). Learn more about missing BASE report.

Additional details and impacted files

@@                    Coverage Diff                     @@
##             eips/amsterdam/eip-8037    #2704   +/-   ##
==========================================================
  Coverage                           ?   88.18%           
==========================================================
  Files                              ?      524           
  Lines                              ?    31120           
  Branches                           ?     3036           
==========================================================
  Hits                               ?    27444           
  Misses                             ?     3161           
  Partials                           ?      515           

Flag	Coverage Δ
unittests	88.18% <ø> (?)

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

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.