Researchers investigate executable stack issues in Linux systems

At the NDSS 2025 conference, Hengkai Ye and Hong Hu from The Pennsylvania State University presented a paper on subtle vulnerabilities in Linux systems that reintroduce executable stacks. Their work highlights how developers, including security experts, accidentally disable protections against code injection attacks. The study examines tools and system components to reveal gaps in enforcing write-xor-execute policies.

The paper, titled 'Too Subtle to Notice: Investigating Executable Stack Issues in Linux Systems,' was delivered in Session 6D on Software Security: Vulnerability Detection during the Network and Distributed System Security Symposium (NDSS) 2025. Historically, code injection via buffer overflows was a common attack vector, but the adoption of write-xor-execute (W^X) protections largely prevented writable-and-executable memory, mitigating such threats.

Despite these advancements, the researchers identified recurring instances where software developers inadvertently disable W^X, restoring executable stacks in popular applications. Each case has been patched, yet the pattern persists in modern development. To explore this, Ye and Hu conducted two key investigations. First, they evaluated program-hardening tools, finding that even experienced security developers often miss critical steps to prevent executable stacks. Notably, 11 tools implemented as inlined reference monitors (IRMs) introduced executable stacks into all 'hardened' applications.

Second, the study analyzed W^X enforcement across Linux's compilation toolchain, kernel, and loader. This process demands tight coordination among multiple components, forming a complex chain of trust for stack protection. A frequent oversight occurs when developers, including security researchers, neglect the GNU-stack section in assembly code, unintentionally enabling executable stacks.

The findings outline potential attacker exploitation paths and offer mitigation recommendations to strengthen Linux security practices. NDSS, which promotes practical advancements in network and distributed system security, hosted the presentation to foster better implementation of security technologies.

Relaterte artikler

Illustration of Russian hackers using Linux VMs to hide malware on Windows systems, showing a computer screen with Hyper-V and malware elements in a dark setting.
Bilde generert av AI

Russian hackers use Linux VMs to hide malware on Windows

Rapportert av AI Bilde generert av AI

Pro-Russian hackers known as Curly COMrades are exploiting Microsoft's Hyper-V technology to embed lightweight Alpine Linux virtual machines within compromised Windows systems. This tactic allows them to run custom malware like CurlyShell and CurlCat undetected by traditional endpoint detection tools. The campaign, uncovered by Bitdefender in collaboration with the Georgian CERT, targets organizations in Europe and beyond.

Cyble Research and Intelligence Labs has revealed ShadowHS, a sophisticated fileless framework for post-exploitation on Linux systems. The tool enables stealthy, in-memory operations and long-term access for attackers. It features a weaponized version of hackshell and advanced evasion techniques.

Rapportert av AI

Ongoing exploitation of the React2Shell vulnerability (CVE-2025-55182)—previously detailed in coverage of China-nexus and cybercriminal campaigns—now includes widespread Linux backdoor installations, arbitrary command execution, and large-scale theft of cloud credentials.

A recent article highlights growing frustrations with d-bus, the longstanding interprocess communication system in Linux, proposing hyprwire as a superior alternative. Critics point to security flaws and poor documentation after nearly two decades of use. The discussion underscores the need for modernization in Linux's core infrastructure.

Rapportert av AI

Following the recent approval of Rust as a permanent kernel language at the 2025 Kernel Maintainers Summit, new details emerge on performance benchmarks, ongoing challenges, and distribution rollouts, solidifying its role in addressing security vulnerabilities.

Researchers at Check Point have revealed that VoidLink, a sophisticated Linux malware targeting cloud servers, was largely built by a single developer using AI tools. The framework, which includes over 30 modular plugins for long-term system access, reached 88,000 lines of code in under a week despite plans suggesting a 20-30 week timeline. This development highlights AI's potential to accelerate advanced malware creation.

Rapportert av AI

Building on the 2025 Kernel Maintainers Summit approval, the Linux kernel finalized permanent Rust integration in late 2025, highlighting early successes like the first Rust CVE detection alongside major performance and security updates in kernel 6.19 and 6.18.

 

 

 

Dette nettstedet bruker informasjonskapsler

Vi bruker informasjonskapsler for analyse for å forbedre nettstedet vårt. Les vår personvernerklæring for mer informasjon.
Avvis