MF2-C02: bound ledger values to Solidity uint256/int256 ranges

[philanton]

Summary

Reject ledger states whose scaled values fall outside the Solidity ABI ranges used for signing and onchain calls. A caller could otherwise submit a home_deposit whose raw amount is 2^256 + X; the node would store the inflated offchain balance, but the signed EVM message and onchain calldata would only carry the low 256 bits, allowing the inflated balance to drain the node vault through a normal transfer_send.

• New core.DecimalToUint256 / core.DecimalToInt256 wrap the (now private) scaling helper and enforce [0, 2^256-1] and [-2^255, 2^255-1] respectively.
• Ledger.Validate now takes assetDecimals and bound-checks every scaled field.
• StateAdvancerV1 looks up token decimals from the asset store and passes them in, so inflated state is rejected before the DB write.
• StatePackerV1 promotes contractLedger to package scope, runs a Validate() guard on it before signing, and uses the bounded helpers when converting decimals.
• coreLedgerToContractLedger and the remaining DepositToNode / Withdraw / Approve / native deposit Value paths swap to DecimalToUint256, so onchain submission is gated by the same check.

Defense in depth: every layer that materialises a *big.Int from a ledger decimal now refuses out-of-range values, instead of relying on a single chokepoint.

Test plan

• go vet ./...
• go test ./...
• New unit tests in pkg/core/utils_test.go for DecimalToUint256 / DecimalToInt256 boundaries (max, max+1, negative, scaling-induced overflow).
• pkg/core/state_packer_test.go::TestPackState_RejectsOutOfRangeValues — packer-level regression for both uint256 and int256 fields.
• pkg/core/state_advancer_test.go::TestValidateAdvancement_RejectsOverflowDeposit — home_deposit with amount 2^256 is rejected before DB write.
• pkg/core/state_advancer_test.go::TestValidateAdvancement_RejectsOverflowNetFlow — net flow 2^255 is rejected.
• Existing app_session_v1 deposit-state tests updated to stub the new GetTokenDecimals call now made by the advancer.

🤖 Generated with Claude Code

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 2f5c5fed-acf1-48ce-8224-ca492ab65ebd

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/nitronode-mf2-c02

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[nksazonov]

Good job! A few missing coverage spots in the new test suite worth addressing before landing.

---

Out-of-diff note — pkg/blockchain/evm/locking_client.go:229: The two call sites in this file were migrated to core.DecimalToUint256, but the original decimalToBigInt(amount decimal.Decimal, decimals int32) *big.Int was left in place with no callers. It is now dead code. Remove it to avoid accidental reuse — its different signature (int32 decimals, no error return) makes it visually distinct from the bounded helpers, but the potential for confusion remains.

Out-of-diff note — pkg/blockchain/evm/utils_test.go:139: TestCoreLedgerToContractLedger_Success covers the happy path after coreLedgerToContractLedger was updated to use DecimalToUint256/DecimalToInt256. There is no rejection test. A test asserting that a Ledger with UserBalance = 2^256 causes coreLedgerToContractLedger to return an error containing "uint256" would give the blockchain-layer conversion the same coverage as the core-layer packer and advancer tests.

[philanton]

@nksazonov on the out-of-diff notes:

1. Dropped the dead decimalToBigInt in pkg/blockchain/evm/locking_client.go (plus the now-unused math/big import) in 4426d54.
2. Added TestCoreLedgerToContractLedger_RejectsOutOfRangeValues covering both the uint256 and int256 paths in the same commit.

[ihsraham]

Approved for C-02. The latest head cleanly bounds ledger values before signing/storage and EVM conversion, and the added tests cover the important closure paths. Thanks for tightening this up.