CVE-2026-39457

7.8

FreeBSD · libnv

A stack-based buffer overflow in libnv allows local attackers to elevate privileges by exploiting improper file descriptor set size validation in select(2) calls.

Executive summary

A stack-based buffer overflow vulnerability in the FreeBSD libnv library could allow an authenticated local attacker to achieve privilege escalation.

Vulnerability

The vulnerability exists because libnv fails to verify if a socket descriptor fits within the FD_SETSIZE limit of 1024 when using select(2). An attacker can trigger stack corruption if they can influence the application to allocate large file descriptors, potentially leading to privilege escalation if the application runs with setuid-root privileges.

Business impact

This flaw carries a CVSS score of 7.8, indicating a high severity risk for local environments. Successful exploitation allows a local user to gain unauthorized elevated privileges, which could lead to full system compromise, data theft, or the execution of arbitrary code with root permissions.

Remediation

Immediate Action: Update FreeBSD systems to the patched releases identified in the FreeBSD security advisory SA-26:16.libnv to resolve the underlying buffer overflow.

Proactive Monitoring: Monitor system logs for unusual process behavior or unexpected crashes in setuid-root applications that utilize libnv.

Compensating Controls: Ensure that the principle of least privilege is strictly enforced for local user accounts to minimize the potential for an attacker to manipulate file descriptor allocations.

Exploitation status

Public Exploit Available: Unknown

Analyst recommendation

Given the potential for privilege escalation, this vulnerability poses a significant risk to the integrity and confidentiality of affected FreeBSD systems. Administrators must prioritize applying the vendor-supplied security patches to the relevant FreeBSD releases immediately to eliminate the possibility of stack-based exploitation.

More FreeBSD CVEs

Sources

Originally found and disclosed by Joshua Rogers of AISLE Research Team, per the CVE Program record.