68870 sc insight reward calculation intermediate multiplication overflow

Submitted on Mar 11th 2026 at 18:15:40 UTC by @ZenHunter for Audit Comp | Folks Finance: Staking Contracts

• Report ID: #68870

• Report Type: Smart Contract

• Report severity: Insight

• Target: https://github.com/Folks-Finance/folks-staking-contracts/blob/main/src/Staking.sol

• Impacts:

Description

Brief/Intro

In Staking._stake, the reward is computed by multiplying three terms — amount, aprBps, and stakingDurationSeconds — before dividing by the denominator. In Solidity 0.8.x, this multiplication reverts on overflow. For sufficiently large inputs the intermediate product exceeds type(uint256).max, causing every stake() call for that input combination to revert unexpectedly and permanently denying the staking service to affected users. Math.mulDiv (which uses 512-bit intermediate arithmetic and never overflows) is already imported via OpenZeppelin (import {Math} from "@openzeppelin/contracts/utils/math/Math.sol"), so this is a one-line change with no new dependencies.

Vulnerability Details

The affected computation is at Staking.sol#L267–L268:

// src/Staking.sol#L267-L268
uint256 rewardBpsDenominator = 1e4 * 365 days;
uint256 reward = (amount * stakingPeriod.aprBps * stakingPeriod.stakingDurationSeconds) / rewardBpsDenominator;

The expression evaluates left-to-right:

1. amount * stakingPeriod.aprBps — first intermediate product

2. (result) * stakingPeriod.stakingDurationSeconds — second intermediate product

3. / rewardBpsDenominator — division applied only after both multiplications

In Solidity 0.8.x, step 2 reverts with a panic if the product exceeds type(uint256).max. The division that would reduce the value never executes.

As APR or duration increase, the threshold decreases and becomes reachable at lower amounts. The fix — Math.mulDiv — is already imported in the contract and performs the equivalent (a * b) / c calculation without intermediate overflow.

Impact Details

Impact category: Security Best Practices

References

• Affected lines: Staking.sol#L267–L268

• OpenZeppelin Math.mulDiv: already imported at Staking.sol#L13

Recommendation

Replace the chained multiplication with Math.mulDiv, which computes floor(a × b / denominator) using 512-bit intermediate arithmetic and never overflows:

uint256 rewardBpsDenominator = 1e4 * 365 days;
uint256 reward = Math.mulDiv(amount, stakingPeriod.aprBps * stakingPeriod.stakingDurationSeconds, rewardBpsDenominator);

Math.mulDiv is already imported via OpenZeppelin (import {Math} from "@openzeppelin/contracts/utils/math/Math.sol"), so this is a one-line change with no new dependencies.

Proof of Concept

File: report/poc/test/RewardOverflow.t.sol

// SPDX-License-Identifier: MIT
pragma solidity ^0.8.23;

import {Test, console} from "forge-std/Test.sol";
import {Math} from "@openzeppelin/contracts/utils/math/Math.sol";

contract RewardOverflowTest is Test {
    uint256 constant REWARD_BPS_DENOMINATOR = 1e4 * 365 days; // 315_360_000_000

    uint32  constant APR_BPS  = 10_000;   // 100% APR
    uint64  constant DURATION = 365 days; // 31_536_000 s

    // One unit above the safe threshold — overflows.
    uint256 constant OVERFLOW_AMOUNT = type(uint256).max / (uint256(APR_BPS) * DURATION) + 1;

    // External wrapper so vm.expectRevert can catch the arithmetic panic.
    function rewardCurrent(uint256 amount) external pure returns (uint256) {
        // mirrors Staking.sol#L267-L268 exactly
        return (amount * APR_BPS * DURATION) / REWARD_BPS_DENOMINATOR;
    }

    function test_Comparison() public {
        console.log("amount              :", OVERFLOW_AMOUNT);

        // --- current: panics ---
        vm.expectRevert(); // arithmetic overflow (panic 0x11)
        this.rewardCurrent(OVERFLOW_AMOUNT);
        console.log("current code        : REVERT (arithmetic overflow)");

        // --- OZ Math.mulDiv: succeeds ---
        uint256 reward = Math.mulDiv(OVERFLOW_AMOUNT, uint256(APR_BPS) * DURATION, REWARD_BPS_DENOMINATOR);
        console.log("Math.mulDiv result  :", reward);
    }
}

Command:

forge test --match-path "test/RewardOverflow.t.sol" -vv

Output:

Ran 1 test for test/RewardOverflow.t.sol:RewardOverflowTest
[PASS] test_Comparison() (gas: 13579)
Logs:
  amount              : 367174306308080274681541682549111833629090514540970839800410908194
  current code        : REVERT (arithmetic overflow)
  Math.mulDiv result  : 367174306308080274681541682549111833629090514540970839800410908194

Suite result: ok. 1 passed; 0 failed; 0 skipped; finished in 1.62ms