source stringclasses 1
value | protocol stringclasses 111
values | report_title stringclasses 111
values | report_date stringdate 2023-03-13 00:00:00 2023-04-11 00:00:00 ⌀ | finding_id stringclasses 150
values | severity stringclasses 6
values | title stringlengths 10 284 | full_content stringlengths 103 34.2k | description stringlengths 46 14.4k ⌀ | proof_of_concept stringclasses 438
values | impact stringlengths 0 4.93k ⌀ | recommended_mitigation stringlengths 2 10.6k ⌀ | code_block_count int64 0 32 | has_poc bool 2
classes | has_mitigation bool 2
classes | content_length int64 103 34.2k | severity_normalized stringclasses 5
values | severity_raw stringclasses 6
values | category stringclasses 16
values | primary_code stringlengths 0 15k | primary_code_language stringclasses 11
values | all_code_blocks stringlengths 0 20.8k | all_code_blocks_count int64 0 32 | solidity_code stringlengths 0 19.9k | has_impact bool 2
classes | has_code_in_description bool 2
classes | content_richness_score float64 0 10.5 | quality_estimate stringclasses 3
values |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | H-01 | high | Attackers can steal tokens and break the protocol's invariant | Attackers can steal tokens and break the protocol's invariant
### Description
The protocol exposes an external function `Well::swapFrom()` which allows any caller to swap `fromToken` to `toToken`.
The function `Well::_getIJ()` is used to get the index of the `fromToken` and `toToken` in the Well's `tokens`.
But the f... | The protocol exposes an external function `Well::swapFrom()` which allows any caller to swap `fromToken` to `toToken`.
The function `Well::_getIJ()` is used to get the index of the `fromToken` and `toToken` in the Well's `tokens`.
But the function `Well::_getIJ` is not implemented correctly.
```solidity
Well.sol
566: ... | Assume a Well with two tokens `t0, t1` is deployed with `ConstantProduct2.sol` as the Well function.
1. The protocol is in a state of `(400 ether, 100 ether)` (`reserve0, reserve1`).
2. An attacker Alice calls `swapFrom(t1, t1, 100 ether, 0)`.
3. At [L148](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb... | The protocol aims for a generalized constant function AMM (CFAMM) and the core invariant of the protocol is there are always more reserved tokens than the actual token balance (`reserves[i] >= tokens[i].balanceOf(well) for all i`).
The incorrect implementation of `_getIJ()` allows attackers to break this invariant and ... | - Add a sanity check to revert if `fromToken==toToken` in the function `Well::swapFrom()` and `Well::swapTo()` .
- Add a sanity check to revert if `iToken==jToken` in the function `Well::_getIJ()` assuming this internal function is not supposed to used with same tokens.
- We strongly recommend adding a check in the fun... | 3 | true | true | 6,248 | high | high | token | Well.sol
566: function _getIJ(//@audit returns (i, 0) if iToken==jToken while it should return (i, i)
567: IERC20[] memory _tokens,
568: IERC20 iToken,
569: IERC20 jToken
570: ) internal pure returns (uint i, uint j) {
571: for (uint k; k < _tokens.length; ++k) {
572: ... | solidity | // Code block 1 (solidity):
Well.sol
566: function _getIJ(//@audit returns (i, 0) if iToken==jToken while it should return (i, i)
567: IERC20[] memory _tokens,
568: IERC20 iToken,
569: IERC20 jToken
570: ) internal pure returns (uint i, uint j) {
571: for (uint k; k < _tokens.len... | 3 | Well.sol
566: function _getIJ(//@audit returns (i, 0) if iToken==jToken while it should return (i, i)
567: IERC20[] memory _tokens,
568: IERC20 iToken,
569: IERC20 jToken
570: ) internal pure returns (uint i, uint j) {
571: for (uint k; k < _tokens.length; ++k) {
572: ... | true | true | 9.6 | high |
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | H-02 | high | Attacker can steal reserves and subsequent liquidity deposits due to lack of input token validation | Attacker can steal reserves and subsequent liquidity deposits due to lack of input token validation
### Description
The protocol exposes an external function `Well::swapFrom()` which allows any caller to swap `fromToken` to `toToken`.
If one of the parameters `fromToken/toToken` is not in `_tokens`, this causes simil... | The protocol exposes an external function `Well::swapFrom()` which allows any caller to swap `fromToken` to `toToken`.
If one of the parameters `fromToken/toToken` is not in `_tokens`, this causes similar issues in `_getIJ` with an index `i/j` being returned as zero.
It appears you can specify a garbage `fromToken` to ... | Beliw is a test case to show this exploit scenario.
The attacker can deploy his own garbage token and call `Well::swapFrom(garbageToken, tokens[1])` that drains the `tokens[0]` balance of the Well.
Note that the similar exploit is also possible for `Well::swapTo()`.
```solidity
function test_exploitGarbageFromToken() ... | The insufficient sanity check on the input tokens of `swapFrom()`(and `swapTo()`) allows attackers to extract tokens and break the protocol's invariant.
Because this exploit does not require any additional assumptions, we evaluate the severity to HIGH. | - Add a sanity check to revert if either `iToken` or `jToken` is not found in the `_tokens` array.
- We also strongly recommend adding a check in the function `Well::_executeSwap()` to make sure the Well has enough reserves on every transaction. | 2 | true | true | 4,727 | high | high | input-validation | function test_exploitGarbageFromToken() prank(user) public {
// this is the maximum that can be sent to the well before hitting ByteStorage: too large
uint256 inAmount = type(uint128).max - tokens[0].balanceOf(address(well));
IERC20 garbageToken = IERC20(new MockToken("GarbageToken", "GTKN", 18));
Mock... | solidity | // Code block 1 (solidity):
function test_exploitGarbageFromToken() prank(user) public {
// this is the maximum that can be sent to the well before hitting ByteStorage: too large
uint256 inAmount = type(uint128).max - tokens[0].balanceOf(address(well));
IERC20 garbageToken = IERC20(new MockToken("GarbageTo... | 2 | function test_exploitGarbageFromToken() prank(user) public {
// this is the maximum that can be sent to the well before hitting ByteStorage: too large
uint256 inAmount = type(uint128).max - tokens[0].balanceOf(address(well));
IERC20 garbageToken = IERC20(new MockToken("GarbageToken", "GTKN", 18));
Mock... | true | true | 7.76 | high |
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | H-03 | high | `removeLiquidity` logic is not correct for general Well functions other than ConstantProduct | `removeLiquidity` logic is not correct for general Well functions other than ConstantProduct
### Description
The protocol aims for a generalized permission-less CFAMM (constant function AMM) where various Well functions can be used.
At the moment, only constant product Well function types are defined but we assume s... | The protocol aims for a generalized permission-less CFAMM (constant function AMM) where various Well functions can be used.
At the moment, only constant product Well function types are defined but we assume support for more generalized Well functions are intended.
The current implementation of `removeLiquidity()` and... | We wrote a test case with the quadratic Well function used by Numoen.
```solidity
// QuadraticWell.sol
/**
* SPDX-License-Identifier: MIT
**/
pragma solidity ^0.8.17;
import "src/interfaces/IWellFunction.sol";
import "src/libraries/LibMath.sol";
contract QuadraticWell is IWellFunction {
using LibMath for uin... | The current `removeLiquidity()` logic assumes specific conditions on the Well function (specifically, some sort of linearity).
This limits the generalization of the protocol, opposed to its original purpose.
Because this will lead to loss of funds for the liquidity providers for general Well functions, we evaluate the ... | We believe that it is not possible to cover all kinds of Well functions without adding some additional functions in the interface `IWellFunction`.
We recommend adding a new function in the `IWellFunction` interface, possibly in the form of `function calcWithdrawFromLp(uint lpTokenToBurn) returns (uint reserve)`.
The o... | 2 | true | true | 8,193 | high | high | access-control | // QuadraticWell.sol
/**
* SPDX-License-Identifier: MIT
**/
pragma solidity ^0.8.17;
import "src/interfaces/IWellFunction.sol";
import "src/libraries/LibMath.sol";
contract QuadraticWell is IWellFunction {
using LibMath for uint;
uint constant PRECISION = 1e18;//@audit-info assume 1:1 upperbound for this... | solidity | // Code block 1 (solidity):
// QuadraticWell.sol
/**
* SPDX-License-Identifier: MIT
**/
pragma solidity ^0.8.17;
import "src/interfaces/IWellFunction.sol";
import "src/libraries/LibMath.sol";
contract QuadraticWell is IWellFunction {
using LibMath for uint;
uint constant PRECISION = 1e18;//@audit-info as... | 2 | // QuadraticWell.sol
/**
* SPDX-License-Identifier: MIT
**/
pragma solidity ^0.8.17;
import "src/interfaces/IWellFunction.sol";
import "src/libraries/LibMath.sol";
contract QuadraticWell is IWellFunction {
using LibMath for uint;
uint constant PRECISION = 1e18;//@audit-info assume 1:1 upperbound for this... | true | true | 9.35 | high |
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | H-04 | high | Read-only reentrancy | Read-only reentrancy
### Description
The current implementation is vulnerable to read-only reentrancy, especially in the function [removeLiquidity](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/Well.sol#L296).
The implementation does not conform to the [CEI pattern](https:/... | The current implementation is vulnerable to read-only reentrancy, especially in the function [removeLiquidity](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/Well.sol#L296).
The implementation does not conform to the [CEI pattern](https://fravoll.github.io/solidity-patterns/ch... | Below is a test case to show the existing read-only reentrancy.
```solidity
// MockCallbackRecipient.sol
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.17;
import {console} from "forge-std/Test.sol";
contract MockCallbackRecipient {
fallback() external payable {
console.log("here");
(bool ... | Although this is not a direct risk to the protocol itself as it is, this can lead to a critical issue in the future.
We evaluate the severity to HIGH. | Implement the CEI pattern in relevant functions.
For example, the function `Well::removeLiquidity` can be modified as below.
```solidity
function removeLiquidity(
uint lpAmountIn,
uint[] calldata minTokenAmountsOut,
address recipient
) external nonReentrant returns (uint[] memory tokenAmountsOut) {
IER... | 3 | true | true | 4,500 | high | high | reentrancy | // MockCallbackRecipient.sol
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.17;
import {console} from "forge-std/Test.sol";
contract MockCallbackRecipient {
fallback() external payable {
console.log("here");
(bool success, bytes memory result) = msg.sender.call(abi.encodeWithSignature("getR... | solidity | // Code block 1 (solidity):
// MockCallbackRecipient.sol
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.17;
import {console} from "forge-std/Test.sol";
contract MockCallbackRecipient {
fallback() external payable {
console.log("here");
(bool success, bytes memory result) = msg.sender.call(a... | 3 | // MockCallbackRecipient.sol
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.17;
import {console} from "forge-std/Test.sol";
contract MockCallbackRecipient {
fallback() external payable {
console.log("here");
(bool success, bytes memory result) = msg.sender.call(abi.encodeWithSignature("getR... | true | true | 9.26 | high |
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | M-01 | medium | Insufficient support for fee-on-transfer ERC20 tokens | Insufficient support for fee-on-transfer ERC20 tokens
### Description
The Well does not rely on the `balanceOf()` function from ERC20 to retrieve current reserve balances.
This is a good design choice.
Reserves values stored in the protocol should be equal to or less than the actual balance.
The current implementati... | The Well does not rely on the `balanceOf()` function from ERC20 to retrieve current reserve balances.
This is a good design choice.
Reserves values stored in the protocol should be equal to or less than the actual balance.
The current implementation assumes `safeTransfer()` will always increase the actual balance equa... | null | Because this vulnerability is dependent on the tokens, we evaluate the severity to MEDIUM. | - If the protocol does not intend to support these kinds of tokens, prevent them by checking the actual balance increase after calling safeTransfer.
- If the protocol wants to support any kind of ERC20 tokens, use a hook method so that the caller can decide the sending amount and check the balance increase amount after... | 0 | false | true | 1,197 | medium | medium | token | 0 | true | false | 3.83 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | M-02 | medium | Some tokens revert on transfer of zero amount | Some tokens revert on transfer of zero amount
### Description
Well protocol intends to be used with various ERC20 tokens.
Some ERC20 tokens revert on transferring zero amount and it is recommended to transfer only when the amount is positive.([Ref](https://github.com/d-xo/weird-erc20#revert-on-zero-value-transfers))
... | Well protocol intends to be used with various ERC20 tokens.
Some ERC20 tokens revert on transferring zero amount and it is recommended to transfer only when the amount is positive.([Ref](https://github.com/d-xo/weird-erc20#revert-on-zero-value-transfers))
In several places, the current implementation does not check the... | null | For some ERC20 tokens, the protocol's important functions (e.g. `removeLiquidity`) would revert and this can lead to insolvency.
We evaluate the severity to MEDIUM. | Check the transfer amount to be positive before calling transfer functions. | 0 | false | true | 983 | medium | medium | token | 0 | true | false | 3.77 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | M-03 | medium | Need to make sure the tokens are unique for ImmutableTokens | Need to make sure the tokens are unique for ImmutableTokens
### Description
The current implementation does not enforce uniqueness in the `_tokens` of `ImmutableTokens`.
Assuming `_tokens[0]=_tokens[1]`.
An honest liquidity provider calls `addLiquidity([1 ether,1 ether], 200 ether, address)`, resulting in the reserv... | The current implementation does not enforce uniqueness in the `_tokens` of `ImmutableTokens`.
Assuming `_tokens[0]=_tokens[1]`.
An honest liquidity provider calls `addLiquidity([1 ether,1 ether], 200 ether, address)`, resulting in the reserves being `(1 ether, 1 ether)`.
At this point, anyone can call the function `sk... | null | Assuming normal liquidity providers are smart enough to check the tokens before sending funds, the likelihood is low, hence we evaluate the severity to MEDIUM. | Enforce uniqueness of the array `_tokens` in `ImmutableTokens`.
This can also be done in the function `ImmutableTokens::getTokenFromList()`. | 0 | false | true | 874 | medium | medium | token | 0 | true | false | 3.41 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | L-01 | low | Incorrect sload in LibBytes | Incorrect sload in LibBytes
### Description
The function `storeUint128` in `LibBytes` intends to pack uint128 `reserves` starting at the given slot but will actually overwrite the final slot if [storing an odd number of reserves](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/sr... | The function `storeUint128` in `LibBytes` intends to pack uint128 `reserves` starting at the given slot but will actually overwrite the final slot if [storing an odd number of reserves](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/libraries/LibBytes.sol#L78). It is currently... | The following test case demonstrates this issue more clearly:
```solidity
// NOTE: Add to LibBytes.t.sol
function test_exploitStoreAndRead() public {
// Write to storage slot to demonstrate overwriting existing values
// In this case, 420 will be stored in the lower 128 bits of the last slot
bytes32 slot =... | Given that assets are not directly at risk, we evaluate the severity to LOW. | Implement the following fix to load the existing value from storage and pack in the lower bits:
```solidity
sload(add(slot, mul(maxI, 32)))
```
 | 2 | true | true | 3,334 | low | low | token | // NOTE: Add to LibBytes.t.sol
function test_exploitStoreAndRead() public {
// Write to storage slot to demonstrate overwriting existing values
// In this case, 420 will be stored in the lower 128 bits of the last slot
bytes32 slot = RESERVES_STORAGE_SLOT;
uint256 maxI = (NUM_RESERVES_MAX - 1) / 2;
... | solidity | // Code block 1 (solidity):
// NOTE: Add to LibBytes.t.sol
function test_exploitStoreAndRead() public {
// Write to storage slot to demonstrate overwriting existing values
// In this case, 420 will be stored in the lower 128 bits of the last slot
bytes32 slot = RESERVES_STORAGE_SLOT;
uint256 maxI = (NUM... | 2 | // NOTE: Add to LibBytes.t.sol
function test_exploitStoreAndRead() public {
// Write to storage slot to demonstrate overwriting existing values
// In this case, 420 will be stored in the lower 128 bits of the last slot
bytes32 slot = RESERVES_STORAGE_SLOT;
uint256 maxI = (NUM_RESERVES_MAX - 1) / 2;
... | true | true | 8.41 | high |
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-01 | informational | Non-standard storage packing | Non-standard storage packing
Per the [Solidity docs](https://docs.soliditylang.org/en/v0.8.17/internals/layout_in_storage.html), the first item in a packed storage slot is stored lower-order aligned; however, [manual packing](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/lib... | null | null | null | null | 0 | false | false | 495 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-02 | informational | EIP-1967 second pre-image best practice | EIP-1967 second pre-image best practice
When calculating custom [EIP-1967](https://eips.ethereum.org/EIPS/eip-1967) storage slots, as in [Well.sol::RESERVES_STORAGE_SLOT](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/Well.sol#L37), it is [best practice](https://ethereum-magic... | null | null | null | null | 0 | false | false | 493 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-03 | informational | Remove experimental ABIEncoderV2 pragma | Remove experimental ABIEncoderV2 pragma
ABIEncoderV2 is enabled by default in Solidity 0.8, so [two](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/interfaces/IWellFunction.sol#L6) [instances](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe... | null | null | null | null | 0 | false | false | 366 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-04 | informational | Inconsistent use of decimal/hex notation in inline assembly | Inconsistent use of decimal/hex notation in inline assembly
For readability and to prevent errors when working with inline assembly, decimal notation should be used for integer constants and hex notation for memory offsets. | null | null | null | null | 0 | false | false | 223 | informational | informational | rounding | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-05 | informational | Unused variables, imports and errors | Unused variables, imports and errors
In `LibBytes`, the [`temp` variable]((https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/libraries/LibBytes.sol#L39)) of `storeUint128` is unused and should be removed.
In `LibMath`:
- OpenZeppelin SafeMath is imported but not used
- `PRBMath_... | null | null | null | null | 0 | false | false | 369 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-06 | informational | Inconsistency in LibMath comments | Inconsistency in LibMath comments
There is inconsistent use of `x` in comments and `a` in code within the `nthRoot` and `sqrt` [functions](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/libraries/LibMath.sol#L44-L147) of `LibMath`. | null | null | null | null | 0 | false | false | 274 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-07 | informational | FIXME and TODO comments | FIXME and TODO comments
There are several [FIXME](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/interfaces/IWell.sol#L268) and [TODO](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/libraries/LibMath.sol#L36) comments that should be a... | null | null | null | null | 0 | false | false | 329 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-08 | informational | Use correct NatSpec tags | Use correct NatSpec tags
Uses of `@dev See {IWell.fn}` should be replaced with `@inheritdoc IWell` to inherit the NatSpec documentation from the interface. | null | null | null | null | 0 | false | false | 155 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-09 | informational | Format for readability | Format for readability
For readability, code should be formatted according to the [Solidity Style Guide](https://docs.soliditylang.org/en/v0.8.17/style-guide.html#other-recommendations) which includes surrounding operators with a single space on either side: e.g. [`numberOfBytes0 - 1`](https://github.com/BeanstalkFarms... | null | null | null | null | 0 | false | false | 408 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | NC-10 | informational | Spelling errors | Spelling errors
The following spelling errors were identified:
- ['configurating'](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/interfaces/IWell.sol#L110) should become 'configuration'
- ['Pump'/'_pumo'](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf9... | null | null | null | null | 0 | false | false | 383 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | G-1 | gas | Simplify modulo operations | Simplify modulo operations
In `LibBytes::storeUint128` and `LibBytes::readUint128`, `reserves.lenth % 2 == 1` and `i % 2 == 1` can be simplified to `reserves.length & 1 == 1` and `i & 1 == 1`. | null | null | null | null | 0 | false | false | 192 | informational | gas | other | 0 | false | false | 0 | low | ||||
cyfrin | 2023-03-13-beanstalk_wells_v0.1 | Beanstalk Wells Initial Audit Report | March 13, 2023 | G-2 | gas | Branchless optimization | Branchless optimization
The `sqrt` function in `MathLib` and [related comment](https://github.com/BeanstalkFarms/Wells/blob/7c498215f843620cb24ec5bbf978c6495f6e5fe4/src/libraries/LibMath.sol#L136-L145) should be updated to reflect changes in Solmate's `FixedPointMathLib` which now includes the [branchless optimization]... | null | null | null | null | 0 | false | false | 485 | informational | gas | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | M-01 | medium | Bad signature validation, malleability & lack of zero address protection in `updateValidatorSet` | Bad signature validation, malleability & lack of zero address protection in `updateValidatorSet`
### Description
The `ecrecover` built-in will return `address(0)` if it fails to recover the signer from a message digest and corresponding signature. There is no `address(0)` check in [`Signature::recoverSigner`](https:/... | The `ecrecover` built-in will return `address(0)` if it fails to recover the signer from a message digest and corresponding signature. There is no `address(0)` check in [`Signature::recoverSigner`](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Signature.sol#L62) or [`Bridge:... | The following forge test can be seen to demonstrate these findings:
```solidity
function test_signatureMalleabilityAndBadValidation() public {
address sender = makeAddr("alice");
address pwner = makeAddr("pwner");
uint256 amount = 1e5;
uint256 nonce = 0;
ValidatorSet memory validatorSet = Validator... | Given the vulnerability described has a low likelihood but high impact, we evaluate the severity to MEDIUM. | As mentioned above, it is recommended to correctly validate recovered signers and protect against signature malleability by using the [OpenZeppelin ECDSA library](https://github.com/OpenZeppelin/openzeppelin-contracts/blob/master/contracts/utils/cryptography/ECDSA.sol). Additionally, add zero address validation to `Bri... | 1 | true | true | 8,406 | medium | medium | input-validation | function test_signatureMalleabilityAndBadValidation() public {
address sender = makeAddr("alice");
address pwner = makeAddr("pwner");
uint256 amount = 1e5;
uint256 nonce = 0;
ValidatorSet memory validatorSet = ValidatorSet(0, s_validators,s_powers);
Signature[] memory sigs = _getSignatures(sende... | solidity | // Code block 1 (solidity):
function test_signatureMalleabilityAndBadValidation() public {
address sender = makeAddr("alice");
address pwner = makeAddr("pwner");
uint256 amount = 1e5;
uint256 nonce = 0;
ValidatorSet memory validatorSet = ValidatorSet(0, s_validators,s_powers);
Signature[] memory... | 1 | function test_signatureMalleabilityAndBadValidation() public {
address sender = makeAddr("alice");
address pwner = makeAddr("pwner");
uint256 amount = 1e5;
uint256 nonce = 0;
ValidatorSet memory validatorSet = ValidatorSet(0, s_validators,s_powers);
Signature[] memory sigs = _getSignatures(sende... | true | true | 8.5 | high |
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | M-02 | medium | Incorrect initialization of `powerThreshold` and lack of validation | Incorrect initialization of `powerThreshold` and lack of validation
### Description
The `Bridge::powerThreshold` should be 2/3 the sum of the validator power across current validators at any point; however, the [calculation](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Br... | The `Bridge::powerThreshold` should be 2/3 the sum of the validator power across current validators at any point; however, the [calculation](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L94) in `Bridge2::constructor` is not correct and it's initialized to `(2 * _... | null | It is possible malicious actions that are lack of validation power are allowed due to wrong initialization of `powerThreshold`.
Because the initializer is assumed to be called by an admin and there is also another admin function to change the `powerThreshold`, we evaluate the severity to MEDIUM. | - Initialize `powerThreshold` to the 2/3 the sum of the initial validator powers.
- Add validation for the new `powerThreshold` in `Bridge::changePowerThreshold()` to allow only reasonable values (ranging from 2/3 to total of the current cumulative validation power). | 0 | false | true | 2,686 | medium | medium | access-control | 0 | true | false | 5.5 | medium | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | L-01 | low | Prevent setting the new epoch to an arbitrary large value | Prevent setting the new epoch to an arbitrary large value
### Description
In the function `Bridge::updateValidatorSet`, the new epoch is [checked](https://github.com/ChainAccelOrg/hyperliquid-audit/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L213) to be greater than the current epoch. However, there is n... | In the function `Bridge::updateValidatorSet`, the new epoch is [checked](https://github.com/ChainAccelOrg/hyperliquid-audit/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L213) to be greater than the current epoch. However, there is no check to prevent setting the new epoch to an arbitrary large value which c... | null | This is not likely to happen but we evaluate as LOW because it can freeze the protocol, locking funds and effectively make the whole protocol insolvent. | It is recommended to allow increase of epoch only to some limited extent, rather than any arbitrary large number. | 0 | false | true | 1,169 | low | low | other | 0 | true | false | 3.35 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-01 | informational | Make `minTotalValidatorPower` immutable | Make `minTotalValidatorPower` immutable
Given that `Bridge::minTotalValidatorPower` is initialized once in the constructor and never modified thereafter, it can be made immutable. This also has the effect of reducing gas usage in functions which currently read its value from storage.
### `f045dbf` Resolution
`minTot... | null | null | null | null | 0 | false | false | 355 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-02 | informational | Avoid using an initializer unless absolutely necessary | Avoid using an initializer unless absolutely necessary
[This comment](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L85-L86) references the potential introduction of a separate initializer but it should be noted that bad initialization is the cause of many exploi... | null | null | null | null | 0 | false | false | 435 | informational | informational | initialization | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-03 | informational | Use calldata for all function arguments that are not modified | Use calldata for all function arguments that are not modified
If function arguments are not intended to be modified then it is best practice to pass them as calldata arguments rather than memory, for example in [https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#LL94C... | null | null | null | null | 0 | false | false | 618 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-04 | informational | `updateValidatorSet` block scope is not necessary | `updateValidatorSet` block scope is not necessary
The block scope in `Bridge::updateValidatorSet` is not necessary as the contract successfully compiles without it and so can be removed.
### `f045dbf` Resolution
The block scope has been removed. | null | null | null | null | 0 | false | false | 248 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-05 | informational | Emit events before interaction | Emit events before interaction
To strictly conform to checks-effect-interactions, it is recommended to emit events prior to any [external interactions](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L122-L123). This is generally advised to ensure correct migration... | null | null | null | null | 0 | false | false | 556 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-06 | informational | Make USDC amount naming more verbose | Make USDC amount naming more verbose
The naming of the USDC amount parameter in [`Bridge::deposit and Bridge::withdraw`](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L120-L152) is not clear, i.e. change `uint256 usdc` to `uin256 usdcAmount`.
### `f045dbf` Resol... | null | null | null | null | 0 | false | false | 559 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-07 | informational | Rename file to `Bridge2.sol` to match the contract name | Rename file to `Bridge2.sol` to match the contract name
The `Bridge2` contract resides in `Bridge.sol`; however, it is best practice to follow the convention of one contract per file with the same name.
### `f045dbf` Resolution
The client team has renamed the contract to `Bridge.sol` with the following comment:
> W... | null | null | null | null | 0 | false | false | 438 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-08 | informational | Add missing NatSpec comments to document function parameters and behaviour | Add missing NatSpec comments to document function parameters and behaviour
It is recommended to document all function behaviours and parameters, especially if they are public-facing.
### `f045dbf` Resolution
The client team acknowledges this finding with the following comment:
> Light comments were added to functio... | null | null | null | null | 0 | false | false | 339 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-09 | informational | No need to explicitly initialize variables to 0 | No need to explicitly initialize variables to 0
Variables are initalized to 0 by default in Solidity, so a [number](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L169-L170) of [superfluous](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c0945070... | null | null | null | null | 0 | false | false | 566 | informational | informational | initialization | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-10 | informational | Use addition assignment operator | Use addition assignment operator
The addition assignment operator `+=` can be used when [checking new validator powers](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Bridge.sol#L238).
### `f045dbf` Resolution
The addition assignment operator has been used. | null | null | null | null | 0 | false | false | 304 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-11 | informational | Localhost chain id can be 1337 or 31337 | Localhost chain id can be 1337 or 31337
The default localhost chain id for some frameworks (e.g. HardHat) is `31337` rather than `1337`. For example, it is not possible to run signature tests using forge without changing `Signature::LOCALHOST_CHAIN_ID` to `31337`.
### `f045dbf` Resolution
This has been resolved by t... | null | null | null | null | 0 | false | false | 343 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-12 | informational | Rename to DOMAIN_SEPARATOR and use `block.chainId` directly | Rename to DOMAIN_SEPARATOR and use `block.chainId` directly
References to `*_DOMAIN_HASH` in `Signature` should be renamed to `*_DOMAIN_SEPARATOR` to be more consistent with the EIP and avoid confusion. Additionally, it is recommended to use `block.chainId` directly, caching on contract creation and only recomputing t... | null | null | null | null | 0 | false | false | 782 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-13 | informational | Rename to `EIP712_DOMAIN_TYPEHASH` | Rename to `EIP712_DOMAIN_TYPEHASH`
Add an [additional underscore](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Signature.sol#L19) for readability.
### `f045dbf` Resolution
The constant has been renamed to `EIP712_DOMAIN_TYPEHASH`. | null | null | null | null | 0 | false | false | 279 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-14 | informational | Include `byte32 salt` in domain typehash | Include `byte32 salt` in domain typehash
Per the [EIP](https://eips.ethereum.org/EIPS/eip-712), `bytes32 salt` should be added to the [domain separator](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Signature.sol#L20) as a last-resort means to distinguish this application f... | null | null | null | null | 0 | false | false | 934 | informational | informational | other | 0 | false | false | 0 | low | ||||
cyfrin | hyperliquid-dex-report | Cyfrin Hyperliquid Audit Report | April 11, 2023 | NC-15 | informational | Verifying contract TODO | Verifying contract TODO
Update the [verifying contract address](https://github.com/hyperliquid-dex/contracts/blob/e0aff464865aa98c09450702d7fb36b1fcd4508c/Signature.sol#L25).
### `f045dbf` Resolution
The client team acknowledges this finding with the following comment:
> Removed the TODO. Keeping the verifying addr... | null | null | null | null | 0 | false | false | 431 | informational | informational | other | 0 | false | false | 0 | low |
End of preview. Expand in Data Studio
No dataset card yet
- Downloads last month
- 7