clet.xyz

The planned Glamsterdam fork on Ethereum’s Hoodi testnet has been indefinitely delayed following a critical report from blockchain security firm ChainSecurity, which flagged significant compatibility risks associated with Ethereum Improvement Proposals (EIP)-8037 and EIP-8038. This unexpected pause in the upgrade rollout has prompted Ethereum core developers to initiate an urgent review of the upgrade specification, scrutinizing the interaction of these EIPs with existing smart contracts. ChainSecurity’s findings are particularly concerning as they stem from an analysis of 13 real-world contract cases, underscoring a practical and immediate threat to the network’s stability if unaddressed.

The Genesis of the Delay: Unforeseen Gas Invariants

The delay in the Glamsterdam fork, a crucial testing phase for upcoming Ethereum upgrades, was officially shared by Ethereum researcher Toni Wahrstätter on X (formerly Twitter) after core developers halted the rollout last week. The Glamsterdam fork was conceived as a vital preparatory step on the Hoodi testnet, designed to stress-test new features and ensure seamless integration before deployment on more prominent testnets and eventually the mainnet. The initial plans for this fork were part of Ethereum’s continuous evolution, aiming to introduce efficiencies and address network complexities.

However, these plans were disrupted by ChainSecurity’s comprehensive report, which highlighted how EIP-8037 specifically interacts with certain contract patterns in an unexpected and potentially detrimental way. Developers dedicated the subsequent weekend to thoroughly assess the scope and nature of these issues, leading to a much clearer understanding of the potential impact.

It has long been understood that contracts which hardcode forwarded gas values could be affected by repricings, where operations like state writes through SSTORE become more expensive under new gas rules. This class of issues is generally anticipated and manageable through developer awareness and updates. The latest findings, however, present a different and more fundamental problem: certain coding patterns within existing smart contracts rely on an invariant—a stable condition or assumption—that EIP-8037 inadvertently breaks.

As Wahrstätter elaborated in his X post, the core of this new problem lies in the behavior of gas accounting: "Gas already spent can come back mid-tx, so gasleft() can go up." This statement reveals a paradigm shift in how gas is handled during a transaction. Historically, developers have largely operated under the assumption that gasleft() (the amount of gas remaining for the current transaction) would only ever decrease or remain constant as a transaction progressed. Contracts designed with this implicit assumption, particularly those performing checks based on gas consumption or relying on a monotonic decrease in available gas, could exhibit unpredictable behavior or outright fail once EIP-8037 is implemented. The implications are far-reaching, as a significant portion of the decentralized applications (dApps) and smart contracts currently deployed on Ethereum are built upon these foundational gas mechanics.

Chronology of a Critical Discovery

The timeline leading to the Glamsterdam delay underscores the rigorous, albeit sometimes challenging, process of Ethereum development:

• Early Planning: Ethereum core developers announce plans for the Glamsterdam fork, targeting the Hoodi testnet as a preliminary testing ground for EIP-8037 and EIP-8038, among other potential changes. The testnet fork was a critical step in a broader roadmap for network enhancements.
• ChainSecurity Engagement: As part of standard security auditing practices for significant network upgrades, ChainSecurity conducts an in-depth analysis of the proposed EIPs, scrutinizing their interaction with the existing contract ecosystem.
• Last Week (Pre-Delay): ChainSecurity submits its report to Ethereum core developers, detailing compatibility risks, particularly concerning EIP-8037’s impact on gas invariants. The report identifies 13 specific real-world contract cases.
• Immediate Developer Response: Upon receiving the report, core developers make the swift decision to pause the Glamsterdam fork rollout on the Hoodi testnet to prevent potential disruptions and allow for thorough investigation.
• Weekend Assessment: Developers dedicate the weekend to meticulously assessing the scope and nature of the issues highlighted by ChainSecurity, gaining a comprehensive understanding of the problem.
• Official Announcement: Ethereum researcher Toni Wahrstätter publicly shares the update on X, explaining the delay and the technical issues identified, clarifying that the picture is now "well understood."
• Current Status: The Glamsterdam fork remains on hold. Core developers and researchers are actively reviewing the upgrade specification and the affected contract patterns, with a focus on deciding the most appropriate path forward.

This sequence highlights the proactive approach of the Ethereum community in prioritizing network integrity and security above rigid adherence to timelines.

The Technical Nuances: Three Classes of Gas-Related Problems

Wahrstätter further delineated the specific technical problems arising from the broken invariant, categorizing them into three main areas, two directly tied to the "gasleft() can go up" scenario, and a third related to gas measurement:

1. Underflow and Transaction Reversion: In the first scenario, contracts that calculate gas usage by comparing a "before" and "after" gas counter within a transaction may experience an underflow. If gas is unexpectedly returned mid-transaction, the "after" value could be higher than the "before" value in certain contexts, leading to an arithmetic underflow and causing the entire transaction to revert. This results in a failed transaction, potentially locking funds or preventing critical dApp functionalities.
2. Unreachable Path Execution: The second problem arises when a sub-call within a transaction fails, and gas is subsequently returned. Under the new EIP-8037 rules, this returned gas could then enable an otherwise "unreachable" execution path within the smart contract. The dangerous aspect here is that the transaction might appear to have succeeded from an external perspective, even though an unintended and potentially malicious code path was executed, leading to subtle but severe vulnerabilities or incorrect state changes. Both outcomes fundamentally stem from the altered mechanism of gas return during a transaction’s lifecycle.
3. Inaccurate Gas Measurement: A separate but related issue concerns contracts and tools that measure gas usage by comparing the gas available before and after a call. EIP-8037 introduces complexities where the "state gas reservoir" – a pool of gas associated with certain operations – is not fully visible or accounted for during execution. Consequently, existing tools and contracts that track gas consumption in this manner might report figures lower than the actual usage, leading to miscalculations, incorrect billing, or flawed resource management within dApps. This discrepancy could undermine the economic models of certain protocols or lead to inefficiencies.

Context: The Importance of Testnets and EIPs in Ethereum’s Evolution

To fully appreciate the significance of this delay, it’s crucial to understand the role of testnets and Ethereum Improvement Proposals (EIPs) in the network’s development lifecycle.

• Testnets (Hoodi, Sepolia, Goerli): Ethereum maintains several testnets, which are parallel blockchains designed to mimic the mainnet environment without using real economic value. Hoodi, in this context, serves as an early-stage testnet, a sandbox where developers can experiment with new features and protocol changes before they are rolled out to larger, more stable testnets like Sepolia or Goerli, and ultimately the Ethereum mainnet. The Glamsterdam fork on Hoodi was intended to be one such early-stage testing event, allowing for the identification of bugs and unforeseen interactions in a controlled environment. The very purpose of these testnets is to catch such critical issues before they can impact real users and billions of dollars in value.
• Ethereum Improvement Proposals (EIPs): EIPs are formal design documents that propose standards for the Ethereum platform, including core protocol specifications, client APIs, and contract standards. EIP-8037 and EIP-8038 are specific proposals aiming to optimize certain aspects of gas economics or transaction processing. Their integration is part of Ethereum’s ongoing efforts to enhance scalability, efficiency, and security. The meticulous review process, including security audits by firms like ChainSecurity, is standard practice for any EIP slated for inclusion in a network upgrade, reflecting the high stakes involved in modifying a global, decentralized financial system.

ChainSecurity’s Role and the Weight of Their Findings

ChainSecurity is a well-regarded blockchain security firm known for its expertise in smart contract auditing and protocol analysis. Their involvement in reviewing critical EIPs is a testament to the Ethereum community’s commitment to robust security. The fact that their report identified 13 real-world cases, rather than purely theoretical constructs, lends immense weight to their findings. These are not abstract possibilities but concrete instances where existing, deployed contracts could break or behave maliciously under the new EIP rules. This practical implication is why core developers are treating the findings as a serious compatibility concern, necessitating a thorough and deliberate response.

Broader Implications for Ethereum Development and the Ecosystem

The delay of the Glamsterdam fork, while inconvenient, is a critical validation of Ethereum’s conservative and security-first approach to protocol development. It highlights several key implications:

• Prioritization of Security and Stability: The immediate halt demonstrates that the Ethereum community prioritizes the integrity and security of the network above all else, even if it means adjusting timelines. This builds trust within the ecosystem, assuring dApp developers and users that upgrades are rigorously vetted.
• Impact on DApp Developers: The identified issues directly affect dApp developers. Those whose contracts rely on the broken gas invariant will need to understand the implications and potentially refactor their code. The call from Toni Wahrstätter for app developers and auditors to "check your contracts for issues with EIP-8037/EIP-8038, and report them" is a clear signal of the collaborative effort required to mitigate these risks. This could involve significant development work for affected projects.
• Evolution of Gas Economics: The incident underscores the complex and often subtle nature of gas economics on Ethereum. Changes, even seemingly minor ones, can have cascading effects across the vast landscape of deployed smart contracts. This necessitates a deep understanding of historical contract patterns and careful foresight in EIP design.
• Future Upgrade Roadmaps: While this specific delay affects a testnet fork, it inevitably introduces uncertainty into the broader Ethereum upgrade roadmap. Core developers will need to decide whether to modify the EIPs, provide migration paths, or implement other solutions, which could impact the timing of subsequent planned upgrades. However, such delays are a normal part of a robust development process, as seen in past major upgrades like The Merge and Dencun, where meticulous planning and testing often led to adjusted schedules.
• The Value of Auditing: The incident reinforces the indispensable role of independent security auditing firms like ChainSecurity. Their ability to identify complex, real-world vulnerabilities before deployment is invaluable to the long-term health and security of the Ethereum network.

Resolution and Next Steps: A Collaborative Path Forward

Wahrstätter’s statement that the Hoodi delay "bought time but won’t resolve the issue" encapsulates the current situation. The pause provides crucial breathing room for investigation, but the underlying compatibility challenges with EIP-8037 and EIP-8038 remain. This week, core contributors are focused on a thorough examination of affected contract patterns to ensure no potential vulnerability is overlooked.

The path to resolution could take several forms, ranging from extensive coordination to fundamental specification changes:

• Ecosystem-Wide Coordination: The simplest resolution, if feasible, would involve a concerted effort among dApp developers, core developers, auditors, and node operators. This might entail providing clear guidelines for contract adjustments, issuing warnings, and facilitating migrations for affected dApps. This approach would minimize changes to the EIPs themselves but place a burden on the application layer.
• Modification of EIP-8037/EIP-8038: In the worst-case scenario, if the issues are deemed too fundamental or widespread to be addressed through coordination alone, the specification of EIP-8037 or EIP-8038 itself might need to be revised. This could involve redesigning aspects of how gas is accounted for or returned to preserve the critical invariant that existing contracts rely upon. Such changes would naturally add more time to the development cycle.
• Call to Action: Wahrstätter’s direct appeal to app developers and auditors to "check your contracts for issues with EIP-8037/EIP-8038, and report them" is a vital call to action. It encourages the wider Ethereum ecosystem to participate in the problem-solving process, leveraging collective intelligence to identify the full scope of the problem and contribute to potential solutions. This decentralized approach to security and development is a hallmark of the Ethereum community.

The ongoing review underscores the dynamic and collaborative nature of Ethereum protocol development. While the delay of the Glamsterdam fork is a temporary setback, it serves as a powerful reminder of the rigorous security standards and the meticulous care taken to ensure the long-term stability and integrity of the world’s leading decentralized computing platform. The ultimate goal remains to deliver a robust and secure network for all users and developers, even if it means adjusting timelines along the way.