chore(deps): update dependency @openzeppelin/contracts to v4.9.6 [security]

[renovate]

This PR contains the following updates:

Package	Change	Age	Adoption	Passing	Confidence
@openzeppelin/contracts (source)	4.9.0 -> 4.9.6

GitHub Vulnerability Alerts

CVE-2023-34234

Impact

By frontrunning the creation of a proposal, an attacker can become the proposer and gain the ability to cancel it. The attacker can do this repeatedly to try to prevent a proposal from being proposed at all.

This impacts the Governor contract in v4.9.0 only, and the GovernorCompatibilityBravo contract since v4.3.0.

Patches

The problem has been patched in 4.9.1 by introducing opt-in frontrunning protection.

Workarounds

Submit the proposal creation transaction to an endpoint with frontrunning protection.

Credit

Reported by Lior Abadi and Joaquin Pereyra from Coinspect.

References

https://www.coinspect.com/openzeppelin-governor-dos/

CVE-2023-34459

Impact

When the verifyMultiProof, verifyMultiProofCalldata, processMultiProof, or processMultiProofCalldata functions are in use, it is possible to construct merkle trees that allow forging a valid multiproof for an arbitrary set of leaves.

A contract may be vulnerable if it uses multiproofs for verification and the merkle tree that is processed includes a node with value 0 at depth 1 (just under the root). This could happen inadvertently for balanced trees with 3 leaves or less, if the leaves are not hashed. This could happen deliberately if a malicious tree builder includes such a node in the tree.

A contract is not vulnerable if it uses single-leaf proving (verify, verifyCalldata, processProof, or processProofCalldata), or if it uses multiproofs with a known tree that has hashed leaves. Standard merkle trees produced or validated with the @​openzeppelin/merkle-tree library are safe.

Patches

The problem has been patched in 4.9.2.

Workarounds

If you are using multiproofs: When constructing merkle trees hash the leaves and do not insert empty nodes in your trees. Using the @​openzeppelin/merkle-tree package eliminates this issue. Do not accept user-provided merkle roots without reconstructing at least the first level of the tree. Verify the merkle tree structure by reconstructing it from the leaves.

CVE-2023-40014

Impact

OpenZeppelin Contracts is a library for secure smart contract development. Starting in version 4.0.0 and prior to version 4.9.3, contracts using ERC2771Context along with a custom trusted forwarder may see _msgSender return address(0) in calls that originate from the forwarder with calldata shorter than 20 bytes. This combination of circumstances does not appear to be common, in particular it is not the case for MinimalForwarder from OpenZeppelin Contracts, or any deployed forwarder the team is aware of, given that the signer address is appended to all calls that originate from these forwarders.

Patches

The problem has been patched in v4.9.3.

CVE-2024-27094

Impact

The Base64.encode function encodes a bytes input by iterating over it in chunks of 3 bytes. When this input is not a multiple of 3, the last iteration may read parts of the memory that are beyond the input buffer.

Although the encode function pads the output for these cases, up to 4 bits of data are kept between the encoding and padding, corrupting the output if these bits were dirty (i.e. memory after the input is not 0). These conditions are more frequent in the following scenarios:

• A bytes memory struct is allocated just after the input and the first bytes of it are non-zero.
• The memory pointer is set to a non-empty memory location before allocating the input.

Developers should evaluate whether the extra bits can be maliciously manipulated by an attacker.

Patches

Upgrade to 5.0.2 or 4.9.6.

References

This issue was reported by the Independent Security Researcher Riley Holterhus through Immunefi (@​rileyholterhus on X)

---

Release Notes

OpenZeppelin/openzeppelin-contracts (@​openzeppelin/contracts)

v4.9.6

Compare Source

• Base64: Fix issue where dirty memory located just after the input buffer is affecting the result. (#​4929)

v4.9.5

Compare Source

• Multicall: Make aware of non-canonical context (i.e. msg.sender is not _msgSender()), allowing compatibility with ERC2771Context. Patch duplicated Address.functionDelegateCall in v4.9.4 (removed).

v4.9.4

Compare Source

• ERC2771Context and Context: Introduce a _contextPrefixLength() getter, used to trim extra information appended to msg.data.
• Multicall: Make aware of non-canonical context (i.e. msg.sender is not _msgSender()), allowing compatibility with ERC2771Context.

v4.9.3

Compare Source

Note
This release contains a fix for GHSA-g4vp-m682-qqmp.

• ERC2771Context: Return the forwarder address whenever the msg.data of a call originating from a trusted forwarder is not long enough to contain the request signer address (i.e. msg.data.length is less than 20 bytes), as specified by ERC-2771. (#​4481)
• ERC2771Context: Prevent revert in _msgData() when a call originating from a trusted forwarder is not long enough to contain the request signer address (i.e. msg.data.length is less than 20 bytes). Return the full calldata in that case. (#​4484)

v4.9.2

Compare Source

• MerkleProof: Fix a bug in processMultiProof and processMultiProofCalldata that allows proving arbitrary leaves if the tree contains a node with value 0 at depth 1.

v4.9.1

Compare Source

• Governor: Add a mechanism to restrict the address of the proposer using a suffix in the description.

---

Configuration

📅 Schedule: Branch creation - "" (UTC), Automerge - At any time (no schedule defined).

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.