Machine learning model repositories, such as the Hugging Face Model Hub, facilitate model exchanges. However, bad actors can deliver malware through compromised models. Existing defenses, such as safer model formats, restrictive (but inflexible) loading policies, and model scanners, have shortcomings: 44.9% of popular models on Hugging Face still use the insecure pickle format, 15% of these cannot be loaded by restrictive loading policies, and model scanners have both false positives and false negatives. Pickle remains the de facto standard for model exchange, and the ML community lacks a tool that offers transparent safe loading.

Hardening OS kernels against memory errors is generally addressed by protecting security-critical data against corruption and disclosure. However, establishing a sound model for identifying sensitive memory objects in need of protection is hard, leading to emergent attack vectors that can be abused by attackers. In this paper, we propose rethinking how OS kernels are hardened by introducing IUBIK for compartmentalizing kernel memory. IUBIK prevents kernel exploitation by segregating attacker-controlled data-frequently used to manipulate security-critical data-in shadow memory, preventing it from interacting with sensitive kernel objects. To achieve this, IUBIK uses MTE: a recent hardware feature, available in ARM CPUs, which allows mitigating exploits based on both spatial and temporal memory-errors, efficiently. We ensure that segregated objects do not contain sensitive fields, such as pointers, by rewriting their struct definitions. Moreover, we develop a profiling framework that explores the kernel codebase in-depth and records code sites where attacker-controlled objects are allocated, allowing IUBIK to isolate them; our profiler recorded 292 privileged and 212 non-privileged allocation sites for a diverse set of workloads. Finally, we evaluate an implementation of IUBIK for the Linux kernel, across a suite of micro- and macro-benchmarks, demonstrating that our prototype incurs no runtime overhead in most tests and negligible additional memory consumption.

Kernel memory allocators maintain several metadata structures optimized for efficiently managing system memory. However, existing implementations adopt either weak or no protection at all to ensure the integrity of said metadata in the presence of memory errors. In this paper, we first demonstrate how existing memory hardening schemes fall short against several in-kernel memory corruption scenarios. We then present ISLAB: a set of novel (slab-based) heap hardening techniques that aim to ensure the integrity of the memory managed by the kernel, and minimize the incurred runtime, and memory, overhead. ISLAB prevents memory corruption exploits by segregating metadata from within corruptible memory objects into shadow memory. It also relies on a novel SMAP-assisted memory isolation framework, called kSMAP, to protect allocator metadata against adversaries with stronger memory access capabilities. We implemented and evaluated ISLAB atop SLUB, the default slab allocator in Linux, and equipped it with kSMAP to protect process credentials, a popular target in kernel exploitation. Our experiments show that ISLAB incurs no runtime overhead in realistic benchmarks, and moderate overhead in stress tests. Lastly, we show how ISLAB's approach can be generalized to protect the integrity of other kernel subsystems that use corruptible metadata for memory management, such as linked lists.

Historically, researchers have treated memory safety-based and speculative execution attacks as two separate domains. Recent work has introduced Speculative Memory-error Abuse (SMA) attacks, which combine memory corruption vulnerabilities with Spectre-like primitives. Using SMA, an attacker can leak sensitive program information and defeat a wide variety of memory-corruption mitigations, including (K)ASLR, software-based XOM, and even ARM PA, eventually carrying out an end-to-end (architecturally-visible) exploit. We present Eclipse: a novel protection scheme against SMA attacks. Eclipse works by propagating artificial data dependencies onto sensitive data, preventing the CPU from using attacker-controlled data during speculative execution. We demonstrate that Eclipse provides comprehensive protection against speculative-probing and Pacman-style attacks, two prominent examples of Speculative Memory-error Abuse attacks that target both the x86(-64) and ARM architectures. We evaluate the performance of Eclipse on x86-64 and demonstrate that it introduces minimal overhead, compared to alternative hardening approaches, incurring ≈0%--9.5% slowdown on SPEC CPU 2017, up to 8.6% slowdown in real-world applications, and negligible overhead in the Linux kernel.

Restricting dangling pointers from accessing freed memory is a promising technique for mitigating use-after-free vulnerabilities in memory-unsafe programming languages. However, existing solutions suffer from high performance overheads, as they rely on conventional page table manipulation to make dangling pointers inaccessible. In this paper, we present Safeslab: a heap-hardening extension that aims to mitigate use-after-free vulnerabilities via a novel and efficient address aliasing approach. Safeslab assigns multiple virtual aliases to each memory page in the system, and manages their access rights via the recently introduced Memory Protection Keys hardware extension, which is designed to provide a fast alternative to page tables for memory management. This allows Safeslab to drastically reduce the number of page table modifications, while blocking dangling pointers efficiently. We integrated Safeslab into the Linux kernel, replacing its default heap allocator (SLUB). The results of our experimental evaluation with real-world benchmarks show that Safeslab incurs a negligible runtime overhead of up to 4% and moderate memory waste.

Modern applications, written in high-level programming languages, enjoy the security benefits of memory and type safety. Unfortunately, even a single memory-unsafe library can wreak havoc on the rest of an otherwise safe application, nullifying all the security guarantees offered by the high-level language and its managed runtime. We perform a study across the Node.js ecosystem to understand the use patterns of binary add-ons. Taking the identified trends into account, we propose a new hybrid permission model aimed at protecting both a binary add-on and its language-specific wrapper. The permission model is applied all around a native add-on and is enforced through a hybrid language-binary scheme that interposes on accesses to sensitive resources from all parts of the native library. We infer the add-on’s permission set automatically over both its binary and JavaScript sides, via a set of novel program analyses. Applied to a wide variety of native add-ons, we show that our framework, BinWrap, reduces access to sensitive resources, defends against real-world exploits, and imposes an overhead that ranges between 0.71%–10.4%.

We present the design, implementation, and evaluation of SysXCHG: a system call (syscall) filtering enforcement mechanism that enables programs to run in accordance with the principle of least privilege. In contrast to the current, hierarchical design of seccomp-BPF, which does not allow a program to run with a different set of allowed syscalls than its descendants, SysXCHG enables applications to run with "tight" syscall filters, uninfluenced by any future-executed (sub-)programs, by allowing filters to be dynamically exchanged at runtime during execve[at]. As a part of SysXCHG, we also present xfilter: a mechanism for fast filtering using a process-specific view of the kernel's syscall table where filtering is performed. In our evaluation of SysXCHG, we found that our filter exchanging design is performant, incurring ≤= 1.71% slowdown on real-world programs in the PaSH benchmark suite, as well as effective, blocking vast amounts of extraneous functionality, including security-critical syscalls, which the current design of seccomp-BPF is unable to.

We present the design, implementation, and evaluation of FineIBT: a CFI enforcement mechanism that improves the precision of hardware-assisted CFI solutions, like Intel IBT, by instrumenting program code to reduce the valid/allowed targets of indirect forward-edge transfers. We study the design of FineIBT on the x86-64 architecture, and implement and evaluate it on Linux and the LLVM toolchain. We designed FineIBT’s instrumentation to be compact, incurring low runtime and memory overheads, and generic, so as to support different CFI policies. Our prototype implementation incurs negligible runtime slowdowns (≈ 0%–1.94% in SPEC CPU2017 and ≈ 0%–1.92% in real-world applications) outperforming Clang-CFI. Lastly, we investigate the effectiveness/security and compatibility of FineIBT using the ConFIRM CFI benchmarking suite, demonstrating that our instrumentation provides complete coverage in the presence of modern software features, while supporting a wide range of CFI policies with the same, predictable performance.

By prioritizing simplicity and portability, least-privilege engineering has been an afterthought in OS design, resulting in monolithic kernels where any exploit leads to total compromise. μSCOPE (“microscope”) addresses this problem by automatically identifying opportunities for least-privilege separation. μSCOPE replaces expert-driven, semi-automated analysis with a general methodology for exploring a continuum of security vs. performance design points by adopting a quantitative and systematic approach to privilege analysis. We apply the μSCOPE methodology to the Linux kernel by (1) instrumenting the entire kernel to gain comprehensive, fine-grained memory access and call activity; (2) mapping these accesses to semantic information; and (3) conducting separability analysis on the kernel using both quantitative privilege and overhead metrics. We discover opportunities for orders of magnitude privilege reduction while predicting relatively low overheads—at 15% mediation overhead, overprivilege in Linux can be reduced up to 99.8%—suggesting fine-grained privilege separation is feasible and laying the groundwork for accelerating real privilege separation.

For comprehensive analysis of all executable code, and fast turn-around time for transformations, it is essential to operate directly on binaries to enable profiling, security hardening, and architectural adaptation. Disassembling binaries is difficult, and prior work relies on a process virtual machine to translate references on the fly or inefficient binary code patching. Our Egalito recompiler leverages metadata present in current stripped x86_64 and ARM64 binaries to generate a complete disassembly, and allows arbitrary modifications that may affect program layout without any constraints from the original binary. We utilize our own layout-agnostic intermediate representation, which is low-level enough to make the regeneration of output code predictable, yet supports a dual high-level representation for sophisticated analysis. We demonstrate nine binary tools including a novel continuous code randomization technique where Egalito transforms itself, and software emulation of the control-flow integrity in upcoming hardware. We evaluated Egalito on a large set of Debian packages, completely analyzing 99.9% of a selection of 867 executables and libraries; a majority of 149 applicable Debian packages pass all tests under Egalito. On SPEC CPU 2006, thanks to our binary optimizations, Egalito actually observes a 1.7% performance speedup.

Attackers leverage memory corruption vulnerabilities to establish primitives for reading from or writing to the address space of a vulnerable process. These primitives form the foundation for code-reuse and data-oriented attacks. While various defenses against the former class of attacks have proven effective, mitigation of the latter remains an open problem. In this paper, we identify various shortcomings of the x86 architecture regarding memory isolation, and leverage virtualization to build an effective defense against data-oriented attacks. Our approach, called xMP, provides (in-guest) selective memory protection primitives that allow VMs to isolate sensitive data in user or kernel space in disjoint xMP domains. We interface the Xen altp2m subsystem with the Linux memory management system, lending VMs the flexibility to define custom policies. Contrary to conventional approaches, xMP takes advantage of virtualization extensions, but after initialization, it does not require any hypervisor intervention. To ensure the integrity of in-kernel management information and pointers to sensitive data within isolated domains, xMP protects pointers with HMACs bound to an immutable context, so that integrity validation succeeds only in the right context. We have applied xMP to protect the page tables and process credentials of the Linux kernel, as well as sensitive data in various user-space applications. Overall, our evaluation shows that xMP introduces minimal overhead for real-world workloads and applications, and offers effective protection against data-oriented attacks.

Developers today have access to an arsenal of toolkits and libraries for rapid application prototyping. However, when an application loads a library, the entirety of that library's code is mapped into the address space, even if only a single function is actually needed. The unused portion is bloat that can negatively impact software defenses by unnecessarily inflating their overhead or increasing their attack surface. Recent work has explored debloating as a way of alleviating the above problems, when source code is available. In this paper, we investigate whether debloating is possible and practical at the binary level. To this end, we present Nibbler: a system that identifies and erases unused functions within shared libraries. Nibbler works in tandem with defenses like continuous code re-randomization and control-flow integrity, enhancing them without incurring additional run-time overhead. We developed and tested a prototype of Nibbler on x86-64 Linux; Nibbler reduces the size of shared libraries and the number of available functions, for real-world binaries and the SPEC CINT2006 suite, by up to 56% and 82%, respectively. We also demonstrate that Nibbler benefits defenses by showing that: (i) it improves the deployability of a continuous re-randomization system for binaries, namely Shuffler, by increasing its efficiency by 20%, and (ii) it improves certain fast, but coarse and context-insensitive control-flow integrity schemes by reducing the number of gadgets reachable through returns and indirect calls by 75% and 49% on average.

Security is particularly important in robotics, as robots can directly perceive and affect the physical world. We describe the results of a scan of the entire IPv4 address space of the Internet for instances of the Robot Operating System (ROS), a widely used robotics software platform. We identified a number of hosts supporting ROS that are exposed to the public Internet, thereby allowing anyone to access robotic sensors and actuators. As a proof of concept, and with the consent of the relevant researchers, we were able to read image sensor information from and actuate a physical robot present in a research lab in an American university. This paper gives an overview of our findings, including our methodology, the geographic distribution of publicly-accessible platforms, the sorts of sensor and actuator data that is available, and the different kinds of robots and sensors that our scan uncovered. Additionally, we offer recommendations on best practices to mitigate these security issues in the future.

The abundance of memory corruption and disclosure vulnerabilities in kernel code necessitates the deployment of hardening techniques to prevent privilege escalation attacks. As stricter memory isolation mechanisms between the kernel and user space become commonplace, attackers increasingly rely on code reuse techniques to exploit kernel vulnerabilities. Contrary to similar attacks in more restrictive settings, as in web browsers, in kernel exploitation, non-privileged local adversaries have great flexibility in abusing memory disclosure vulnerabilities to dynamically discover, or infer, the location of code snippets in order to construct code-reuse payloads. Recent studies have shown that the coupling of code diversification with the enforcement of a “read XOR execute” (R ∧ X) memory safety policy is an effective defense against the exploitation of userland software, but so far this approach has not been applied for the protection of the kernel itself. In this article, we fill this gap by presenting kR ∧ X: a kernel-hardening scheme based on execute-only memory and code diversification. We study a previously unexplored point in the design space, where a hypervisor or a super-privileged component is not required. Implemented mostly as a set of GCC plugins, kR ∧ X is readily applicable to x86 Linux kernels (both 32b and 64b) and can benefit from hardware support (segmentation on x86, MPX on x86-64) to optimize performance. In full protection mode, kR ∧ X incurs a low runtime overhead of 4.04%, which drops to 2.32% when MPX is available, and 1.32% when memory segmentation is in use.

Despite decades of research on software diversification, only address space layout randomization has seen widespread adoption. Code randomization, an effective defense against return-oriented programming exploits, has remained an academic exercise mainly due to i) the lack of a transparent and streamlined deployment model that does not disrupt existing software distribution norms, and ii) the inherent incompatibility of program variants with error reporting, whitelisting, patching, and other operations that rely on code uniformity. In this work we present compiler-assisted code randomization (CCR), a hybrid approach that relies on compiler-rewriter cooperation to enable fast and robust fine-grained code randomization on end-user systems, while maintaining compatibility with existing software distribution models. The main concept behind CCR is to augment binaries with a minimal set of transformation-assisting metadata, which i) facilitate rapid fine-grained code transformation at installation or load time, and ii) form the basis for reversing any applied code transformation when needed, to maintain compatibility with existing mechanisms that rely on referencing the original code. We have implemented a prototype of this approach by extending the LLVM compiler toolchain, and developing a simple binary rewriter that leverages the embedded metadata to generate randomized variants using basic block reordering. The results of our experimental evaluation demonstrate the feasibility and practicality of CCR, as on average it incurs a modest file size increase of 11.46% and a negligible runtime overhead of 0.28%, while it is compatible with link-time optimization and control flow integrity.

The abundance of memory corruption and disclosure vulnerabilities in kernel code necessitates the deployment of hardening techniques to prevent privilege escalation attacks. As more strict memory isolation mechanisms between the kernel and user space, like Intel's SMEP, become commonplace, attackers increasingly rely on code reuse techniques to exploit kernel vulnerabilities. Contrary to similar attacks in more restrictive settings, such as web browsers, in kernel exploitation, non-privileged local adversaries have great flexibility in abusing memory disclosure vulnerabilities to dynamically discover, or infer, the location of certain code snippets and construct code-reuse payloads. Recent studies have shown that the coupling of code diversification with the enforcement of a "read XOR execute" (R^X) memory safety policy is an effective defense against the exploitation of userland software, but so far this approach has not been applied for the protection of the kernel itself.

Many software providers operate crash reporting services to automatically collect crashes from millions of customers and file bug reports. Precisely triaging crashes is necessary and important for software providers because the millions of crashes that may be reported every day are critical in identifying high impact bugs. However, the triaging accuracy of existing systems is limited, as they rely only on the syntactic information of the stack trace at the moment of a crash without analyzing program semantics.

Over the past decade many exploit mitigation techniques have been introduced to defend against memory corruption attacks. W^X, ASLR, and canary-based protections are nowadays widely deployed and considered standard practice. However, despite the fact that these techniques have evolved over time, they still suffer from limitations that enable skilled adversaries to bypass them.

We present a micro-architectural side-channel attack that runs entirely in the browser. In contrast to previous work in this genre, our attack does not require the attacker to install software on the victim's machine; to facilitate the attack, the victim needs only to browse to an untrusted webpage that contains attacker-controlled content. This makes our attack model highly scalable, and extremely relevant and practical to today's Web, as most desktop browsers currently used to access the Internet are affected by such side channel threats. Our attack, which is an extension to the last-level cache attacks of Liu et al., allows a remote adversary to recover information belonging to other processes, users, and even virtual machines running on the same physical host with the victim web browser. We describe the fundamentals behind our attack, and evaluate its performance characteristics. In addition, we show how it can be used to compromise user privacy in a common setting, letting an attacker spy after a victim that uses private browsing. Defending against this side channel is possible, but the required countermeasures can exact an impractical cost on benign uses of the browser.

Dynamic data flow tracking (DFT) is a technique broadly used in a variety of security applications that, unfortunately, exhibits poor performance, preventing its adoption in production systems. We present ShadowReplica, a new and efficient approach for accelerating DFT and other shadow memory-based analyses, by decoupling analysis from execution and utilizing spare CPU cores to run them in parallel. Our approach enables us to run a heavyweight technique, like dynamic taint analysis (DTA), twice as fast, while concurrently consuming fewer CPU cycles than when applying it in-line. DFT is run in parallel by a second shadow thread that is spawned for each application thread, and the two communicate using a shared data structure. We avoid the problems suffered by previous approaches, by introducing an off-line application analysis phase that utilizes both static and dynamic analysis methodologies to generate optimized code for decoupling execution and implementing DFT, while it also minimizes the amount of information that needs to be communicated between the two threads. Furthermore, we use a lock-free ring buffer structure and an N-way buffering scheme to efficiently exchange data between threads and maintain high cache-hit rates on multi-core CPUs. Our evaluation shows that ShadowReplica is on average ~2.3× faster than in-line DFT (~2.75× slowdown over native execution) when running the SPEC CPU2006 benchmark, while similar speed ups were observed with command-line utilities and popular server software. Astoundingly, ShadowReplica also reduces the CPU cycles used up to 30%.

Applications can be logically separated to parts that face different types of threats, or suffer dissimilar exposure to a particular threat because of external events or innate properties of the software. Based on this observation, we propose the virtual partitioning of applications that will allow the selective and targeted application of those protection mechanisms that are most needed on each partition, or manage an application's attack surface by protecting the most exposed partition. We demonstrate the value of our scheme by introducing a methodology to automatically partition software, based on the intrinsic property of user authentication. Our approach is able to automatically determine the point where users authenticate, without access to source code. At runtime, we employ a monitor that utilizes the identified authentication points, as well as events like accessing specific files, to partition execution and adapt defenses by switching between protection mechanisms of varied intensity, such as dynamic taint analysis and instruction-set randomization. We evaluate our approach using seven well-known network applications, including the MySQL database server. Our results indicate that our methodology can accurately discover authentication points. Furthermore, we show that using virtual partitioning to apply costly protection mechanisms can reduce performance overhead by up to 5x, depending on the nature of the application.

Dynamic data flow tracking (DFT) deals with tagging and tracking data of interest as they propagate during program execution. DFT has been repeatedly implemented by a variety of tools for numerous purposes, including protection from zero-day and cross-site scripting attacks, detection and prevention of information leaks, and for the analysis of legitimate and malicious software. We present libdft, a dynamic DFT framework that unlike previous work is at once fast, reusable, and works with commodity software and hardware. libdft provides an API for building DFT-enabled tools that work on unmodified binaries, running on common operating systems and hardware, thus facilitating research and rapid prototyping. We explore different approaches for implementing the low-level aspects of instruction-level data tracking, introduce a more efficient and 64-bit capable shadow memory, and identify (and avoid) the common pitfalls responsible for the excessive performance overhead of previous studies. We evaluate libdft using real applications with large codebases like the Apache and MySQL servers, and the Firefox web browser. We also use a series of benchmarks and utilities to compare libdft with similar systems. Our results indicate that it performs at least as fast, if not faster, than previous solutions, and to the best of our knowledge, we are the first to evaluate the performance overhead of a fast dynamic DFT implementation in such depth. Finally, libdft is freely available as open source software.

We propose a novel trap-based architecture for detecting passive, “silent”, attackers who are eavesdropping on enterprise networks. Motivated by the increasing number of incidents where attackers sniff the local network for interesting information, such as credit card numbers, account credentials, and passwords, we introduce a methodology for building a trap-based network that is designed to maximize the realism of bait-laced traffic. Our proposal relies on a “record, modify, replay” paradigm that can be easily adapted to different networked environments. The primary contributions of our architecture are the ease of automatically injecting large amounts of believable bait, and the integration of different detection mechanisms in the back-end. We demonstrate our methodology in a prototype platform that uses our decoy injection API to dynamically create and dispense network traps on a subset of our campus wireless network. Our network traps consist of several types of monitored passwords, authentication cookies, credit cards and documents containing beacons to alarm when opened. The efficacy of our decoys against a model attack program is also discussed, along with results obtained from experiments in the field. In addition, we present a user study that demonstrates the believability of our decoy traffic, and finally, we provide experimental results to show that our solution causes only negligible interference to ordinary users.

We propose a novel trap-based architecture for enterprise networks that detects "silent" attackers who are eavesdropping network traffic. The primary contributions of our work are the ease of injecting, automatically, large amounts of believable bait, and the integration of various detection mechanisms in the back-end. We demonstrate our methodology in a prototype platform that uses our decoy injection API to dynamically create and dispense network traps on a subset of our campus wireless network. Finally, we present results of a user study that demonstrates the believability of our automatically generated decoy traffic.

In the past few years numerous P2P file-sharing and content distribution systems have been designed, implemented, and evaluated via simulations, real world measurements, and mathematical analysis. Yet, only few of them have stood the test of time and gained wide user acceptance. BitTorrent is the one that holds the lion's share among them and the reasons behind its success have been studied to a great extent with interesting results. Nevertheless, even though P2P content distribution remains one of the most active research areas, little progress has been made towards the study of the BitTorrent protocol (and its variations), in a fully controllable and realistic simulation environment. In this paper we describe and analyze a full-featured and extensible implementation of BitTorrent for the OMNeT++ simulation platform. Moreover, since we aim at realistic simulations, we present our enhancements on a popular conversion tool for practical Internet topologies, as well as our churn generator that is based on the analysis of real BitTorrent traces. Finally, we set forth the results from the evaluation of our prototype implementation regarding resource demands under different simulation scenarios.

One of the most considerable functions in a hospital's infection control program is the surveillance of antibiotic resistance. Several traditional methods used to measure it do not provide adequate and promising results for further analysis. Data mining techniques, such as the association rules, have been used in the past and successfully led to discovering interesting patterns in public health data. In this work, we present the architecture of a novel framework which integrates data from multiple hospitals, discovers association rules, stores them in a data warehouse for future analysis and provides anytime accessibility through an intuitive Web interface. We implemented the proposed architecture as a Web application and evaluated it using data from the WHONET software installed in many Greek hospitals that belong to "the Greek System for Surveillance of Antimicrobial Resistance" network. The contribution of the proposed framework is considered to be a standardized workflow aiming at the integration of data produced by various hospitals into a consistent data warehouse and the use of a mechanism that detects hidden and previously unknown patterns on large datasets, in terms of association rules, which can provide surveillance warnings.

Nowadays wireless networks are becoming increasingly popular in urban areas, university campuses and corporate environments. This, along with the widespread deployment of mobile devices with advanced computational capabilities, creates more opportunities for developing collaborative/social applications. The development of such applications often raises a number of challenges for the developer mostly due to synchronization issues, increased network failures of the wireless environment and limited resource availability. In this paper we present a novel framework that tackles the above problems and releases the developer from the burden of dealing with cross-domain problems. We focus on the development of a certain type of applications, inside the context of ubiquitous gaming, in a small scale peer-to-peer manner.