CVE-2026-58050: libssh2 Heap Buffer Overflow in Public-Key Subsystem (32-bit)

Technical summary

The vulnerability stems from unsafe integer arithmetic in libssh2's public-key subsystem handler. The code reads an attacker-controlled 32-bit attribute count from the SSH server response without validation, then calculates the required allocation size using the formula: num_attrs * sizeof(libssh2_publickey_attribute). On 32-bit platforms, this multiplication can overflow, wrapping around to a much smaller value. The subsequent attribute-parsing loop treats this count as legitimate and writes attribute structures into the undersized heap buffer, causing an out-of-bounds write. The vulnerability is classified as CWE-190 (Integer Overflow or Wraparound) and requires specific conditions on the target architecture and library version, making exploitation dependent on the victim using a 32-bit build of libssh2.

Business impact

Organizations relying on libssh2 for SSH client functionality face potential compromise through malicious SSH servers. This includes embedded systems, appliances, IoT devices, and applications that bundle libssh2 as a dependency. A successful exploit could allow attackers to execute arbitrary code with the privileges of the affected application, or trigger denial of service. The attack surface includes any user or automated system that connects to untrusted or compromised SSH servers. Patching is critical for systems handling sensitive SSH operations, particularly those in networks where server compromise or man-in-the-middle attacks are plausible.

Affected systems

libssh2 versions through 1.11.1 are affected. The practical impact is primarily on 32-bit builds of libssh2; 64-bit builds are not vulnerable due to the absence of integer overflow in the multiplication on larger word sizes. Affected systems include applications and libraries that statically or dynamically link libssh2, particularly those deployed on ARM, x86, MIPS, and other 32-bit architectures. Verify your specific usage: check build configuration, architecture, and libssh2 version in your supply chain.

Exploitability

Exploitation requires the victim to connect to a malicious or compromised SSH server. No user interaction beyond initiating an SSH connection is needed, and the attack works across networks (AV:N). However, the attack complexity is rated as high (AC:H) due to the specific conditions required: the target must run a vulnerable 32-bit build, the SSH server must be under attacker control or compromised, and the exploit must precisely craft the overflow. Public exploit code is not known to exist, and this has not been added to the CISA KEV catalog. Real-world exploitation is possible but requires active server compromise or man-in-the-middle positioning.

Remediation

Update libssh2 to a patched version released after June 28, 2026. Verify the patch version against the vendor's official advisory to confirm the overflow is addressed. For systems unable to update immediately, implement network controls to restrict SSH connections to trusted, verified servers. Consider disabling public-key subsystem support if it is not required. Monitor process execution and memory access patterns for signs of exploitation. On 64-bit platforms, prioritize patching for other potential vulnerabilities, but treat 32-bit deployments as urgent.

Patch guidance

Contact your libssh2 vendor or check the official libssh2 release notes and security advisories for the specific patch version that resolves CVE-2026-58050. Apply updates in a staged manner: first to non-production systems to verify compatibility, then to production. For embedded systems and appliances, check with the device manufacturer for firmware updates that include the patched libssh2. If using libssh2 as a transitive dependency, ensure your build tooling and package manager pull the updated version. Test SSH client functionality after patching, particularly for applications using the public-key subsystem.

Detection guidance

On 32-bit platforms running vulnerable libssh2 versions, monitor for segmentation faults or memory corruption errors in applications during SSH connections to untrusted servers. Enable debug logging in libssh2 where available to detect anomalies in SSH protocol parsing. Check system logs and core dumps for crashes in the public-key subsystem handler. Network intrusion detection systems should flag SSH servers sending unusually large or suspicious attribute count values in public-key subsystem responses. Conduct binary analysis of deployed libssh2 builds to confirm they are 32-bit and identify the version. If feasible, use runtime memory sanitizers (ASan, MSan) in test environments to catch buffer overflows before production deployment.

Why prioritize this

This vulnerability merits immediate attention due to its high CVSS score (7.0), potential for remote code execution on 32-bit systems, and the commonality of libssh2 in embedded and client-side applications. Although 64-bit platforms are unaffected, organizations with mixed architectures or legacy 32-bit deployments face significant risk. The low barrier to triggering the overflow—merely initiating an SSH connection to a malicious server—increases practical risk in networks where server compromise or MITM attacks are credible threats. Patching should be prioritized for edge cases and mobile systems first, then extended to the broader estate.

Risk score, explained

The CVSS 3.1 score of 7.0 (HIGH) reflects a network-accessible attack vector, high potential impact on confidentiality and integrity, and the ability to cause denial of service. The high attack complexity moderates the score somewhat, accounting for the requirement that a 32-bit build be in use and that the attacker control or compromise the SSH server. The absence of privilege escalation or user interaction keeps the score below CRITICAL. The practical risk is significant for 32-bit deployments in untrusted network environments; organizations running only 64-bit builds should still patch to eliminate architectural variants in their supply chain.

Frequently asked questions