skip to main content


Title: ZombieLoad: Cross-Privilege-Boundary Data Sampling
In early 2018, Meltdown first showed how to read arbitrary kernel memory from user space by exploiting side-effects from transient instructions. While this attack has been mitigated through stronger isolation boundaries between user and kernel space, Meltdown inspired an entirely new class of fault-driven transient-execution attacks. Particularly, over the past year, Meltdown-type attacks have been extended to not only leak data from the L1 cache but also from various other microarchitectural structures, including the FPU register file and store buffer. In this paper, we present the ZombieLoad attack which uncovers a novel Meltdown-type effect in the processor’s fill-buffer logic. Our analysis shows that faulting load instructions (i.e., loads that have to be re-issued) may transiently dereference unauthorized destinations previously brought into the fill buffer by the current or a sibling logical CPU. In contrast to concurrent attacks on the fill buffer, we are the first to report data leakage of recently loaded and stored stale values across logical cores even on Meltdown- and MDS-resistant processors. Hence, despite Intel’s claims [36], we show that the hardware fixes in new CPUs are not sufficient. We demonstrate ZombieLoad’s effectiveness in a multitude of practical attack scenarios across CPU privilege rings, OS processes, virtual machines, and SGX enclaves. We discuss both short and long-term mitigation approaches and arrive at the conclusion that disabling hyperthreading is the only possible workaround to prevent at least the most-powerful cross-hyperthread attack scenarios on current processors, as Intel’s software fixes are incomplete.  more » « less
Award ID(s):
1814406
NSF-PAR ID:
10295738
Author(s) / Creator(s):
; ; ; ; ; ;
Date Published:
Journal Name:
USENIX Security 2020
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. null (Ed.)
    Meltdown and Spectre enable arbitrary data leakage from memory via various side channels. Short-term software mitigations for Meltdown are only a temporary solution with a significant performance overhead. Due to hardware fixes, these mitigations are disabled on recent processors. In this paper, we show that Meltdown-like attacks are still possible on recent CPUs which are not vulnerable to Meltdown. We identify two behaviors of the store buffer, a microarchitectural resource to reduce the latency for data stores, that enable powerful attacks. The first behavior, Write Transient Forwarding forwards data from stores to subsequent loads even when the load address differs from that of the store. The second, Store-to-Leak exploits the interaction between the TLB and the store buffer to leak metadata on store addresses. Based on these, we develop multiple attacks and demonstrate data leakage, control flow recovery, and attacks on ASLR. Our paper shows that Meltdown-like attacks are still possible, and software fixes with potentially significant performance overheads are still necessary to ensure proper isolation between the kernel and user space. 
    more » « less
  2. null (Ed.)
    The recent Spectre attack first showed how to inject incorrect branch targets into a victim domain by poisoning microarchitectural branch prediction history. In this paper, we generalize injection-based methodologies to the memory hierarchy by directly injecting incorrect, attacker-controlled values into a victim's transient execution. We propose Load Value Injection (LVI) as an innovative technique to reversely exploit Meltdown-type microarchitectural data leakage. LVI abuses that faulting or assisted loads, executed by a legitimate victim program, may transiently use dummy values or poisoned data from various microarchitectural buffers, before eventually being re-issued by the processor. We show how LVI gadgets allow to expose victim secrets and hijack transient control flow. We practically demonstrate LVI in several proof-of-concept attacks against Intel SGX enclaves, and we discuss implications for traditional user process and kernel isolation. State-of-the-art Meltdown and Spectre defenses, including widespread silicon-level and microcode mitigations, are orthogonal to our novel LVI techniques. LVI drastically widens the spectrum of incorrect transient paths. Fully mitigating our attacks requires serializing the processor pipeline with lfence instructions after possibly every memory load. Additionally and even worse, due to implicit loads, certain instructions have to be blacklisted, including the ubiquitous x86 ret instruction. Intel plans compiler and assembler-based full mitigations that will allow at least SGX enclave programs to remain secure on LVI-vulnerable systems. Depending on the application and optimization strategy, we observe extensive overheads of factor 2 to 19 for prototype implementations of the full mitigation. 
    more » « less
  3. null (Ed.)
    In May 2019, a new class of transient execution attack based on Meltdown called microarchitectural data sampling (MDS), was disclosed. MDS enables adversaries to leak secrets across security domains by collecting data from shared CPU resources such as data cache, fill buffers, and store buffers. These resources may temporarily hold data that belongs to other processes and privileged contexts, which could falsely be forwarded to memory accesses of an adversary. We perform an in-depth analysis of these Meltdown-style attacks using our novel fuzzing-based approach. We introduce an analysis tool, named Transynther, which mutates the basic block of existing Meltdown variants to generate and evaluate new Meltdown subvariants. We apply Transynther to analyze modern CPUs and better understand the root cause of these attacks. As a result, we find new variants of MDS that only target specific memory operations, e.g., fast string copies. Based on our findings, we propose a new attack, named Medusa, which can leak data from implicit write-combining memory operations. Since Medusa only applies to specific operations, it can be used to pinpoint vulnerable targets. In a case study, we apply Medusa to recover the key during the RSA signing operation. We show that Medusa can leak various parts of an RSA key during the base64 decoding stage. Then we build leakage templates and recover full RSA keys by employing lattice-based cryptanalysis techniques. 
    more » « less
  4. Transient execution attacks, also known as speculative execution attacks, have drawn much interest in the last few years as they can cause critical data leakage. Since the first disclosure of Spectre and Meltdown attacks in January 2018, a number of new transient execution attack types have been demonstrated targeting different processors. A transient execution attack consists of two main components: transient execution itself and a covert channel that is used to actually exfiltrate the information.Transient execution is a result of the fundamental features of modern processors that are designed to boost performance and efficiency, while covert channels are unintended information leakage channels that result from temporal and spatial sharing of the micro-architectural components. Given the severity of the transient execution attacks, they have motivated computer architects in both industry and academia to rethink the design of the processors and to propose hardware defenses. To help understand the transient execution attacks, this survey summarizes the phases of the attacks and the security boundaries across which the information is leaked in different attacks.This survey further analyzes the causes of transient execution as well as the different types of covert channels and presents a taxonomy of the attacks based on the causes and types. This survey in addition presents metrics for comparing different aspects of the transient execution attacks and uses them to evaluate the feasibility of the different attacks. This survey especially considers both existing attacks and potential new attacks suggested by our analysis. This survey finishes by discussing different mitigations that have so far been proposed at the micro-architecture level and discusses their benefits and limitations. 
    more » « less
  5. Obeid, Iyad Selesnick (Ed.)
    Electroencephalography (EEG) is a popular clinical monitoring tool used for diagnosing brain-related disorders such as epilepsy [1]. As monitoring EEGs in a critical-care setting is an expensive and tedious task, there is a great interest in developing real-time EEG monitoring tools to improve patient care quality and efficiency [2]. However, clinicians require automatic seizure detection tools that provide decisions with at least 75% sensitivity and less than 1 false alarm (FA) per 24 hours [3]. Some commercial tools recently claim to reach such performance levels, including the Olympic Brainz Monitor [4] and Persyst 14 [5]. In this abstract, we describe our efforts to transform a high-performance offline seizure detection system [3] into a low latency real-time or online seizure detection system. An overview of the system is shown in Figure 1. The main difference between an online versus offline system is that an online system should always be causal and has minimum latency which is often defined by domain experts. The offline system, shown in Figure 2, uses two phases of deep learning models with postprocessing [3]. The channel-based long short term memory (LSTM) model (Phase 1 or P1) processes linear frequency cepstral coefficients (LFCC) [6] features from each EEG channel separately. We use the hypotheses generated by the P1 model and create additional features that carry information about the detected events and their confidence. The P2 model uses these additional features and the LFCC features to learn the temporal and spatial aspects of the EEG signals using a hybrid convolutional neural network (CNN) and LSTM model. Finally, Phase 3 aggregates the results from both P1 and P2 before applying a final postprocessing step. The online system implements Phase 1 by taking advantage of the Linux piping mechanism, multithreading techniques, and multi-core processors. To convert Phase 1 into an online system, we divide the system into five major modules: signal preprocessor, feature extractor, event decoder, postprocessor, and visualizer. The system reads 0.1-second frames from each EEG channel and sends them to the feature extractor and the visualizer. The feature extractor generates LFCC features in real time from the streaming EEG signal. Next, the system computes seizure and background probabilities using a channel-based LSTM model and applies a postprocessor to aggregate the detected events across channels. The system then displays the EEG signal and the decisions simultaneously using a visualization module. The online system uses C++, Python, TensorFlow, and PyQtGraph in its implementation. The online system accepts streamed EEG data sampled at 250 Hz as input. The system begins processing the EEG signal by applying a TCP montage [8]. Depending on the type of the montage, the EEG signal can have either 22 or 20 channels. To enable the online operation, we send 0.1-second (25 samples) length frames from each channel of the streamed EEG signal to the feature extractor and the visualizer. Feature extraction is performed sequentially on each channel. The signal preprocessor writes the sample frames into two streams to facilitate these modules. In the first stream, the feature extractor receives the signals using stdin. In parallel, as a second stream, the visualizer shares a user-defined file with the signal preprocessor. This user-defined file holds raw signal information as a buffer for the visualizer. The signal preprocessor writes into the file while the visualizer reads from it. Reading and writing into the same file poses a challenge. The visualizer can start reading while the signal preprocessor is writing into it. To resolve this issue, we utilize a file locking mechanism in the signal preprocessor and visualizer. Each of the processes temporarily locks the file, performs its operation, releases the lock, and tries to obtain the lock after a waiting period. The file locking mechanism ensures that only one process can access the file by prohibiting other processes from reading or writing while one process is modifying the file [9]. The feature extractor uses circular buffers to save 0.3 seconds or 75 samples from each channel for extracting 0.2-second or 50-sample long center-aligned windows. The module generates 8 absolute LFCC features where the zeroth cepstral coefficient is replaced by a temporal domain energy term. For extracting the rest of the features, three pipelines are used. The differential energy feature is calculated in a 0.9-second absolute feature window with a frame size of 0.1 seconds. The difference between the maximum and minimum temporal energy terms is calculated in this range. Then, the first derivative or the delta features are calculated using another 0.9-second window. Finally, the second derivative or delta-delta features are calculated using a 0.3-second window [6]. The differential energy for the delta-delta features is not included. In total, we extract 26 features from the raw sample windows which add 1.1 seconds of delay to the system. We used the Temple University Hospital Seizure Database (TUSZ) v1.2.1 for developing the online system [10]. The statistics for this dataset are shown in Table 1. A channel-based LSTM model was trained using the features derived from the train set using the online feature extractor module. A window-based normalization technique was applied to those features. In the offline model, we scale features by normalizing using the maximum absolute value of a channel [11] before applying a sliding window approach. Since the online system has access to a limited amount of data, we normalize based on the observed window. The model uses the feature vectors with a frame size of 1 second and a window size of 7 seconds. We evaluated the model using the offline P1 postprocessor to determine the efficacy of the delayed features and the window-based normalization technique. As shown by the results of experiments 1 and 4 in Table 2, these changes give us a comparable performance to the offline model. The online event decoder module utilizes this trained model for computing probabilities for the seizure and background classes. These posteriors are then postprocessed to remove spurious detections. The online postprocessor receives and saves 8 seconds of class posteriors in a buffer for further processing. It applies multiple heuristic filters (e.g., probability threshold) to make an overall decision by combining events across the channels. These filters evaluate the average confidence, the duration of a seizure, and the channels where the seizures were observed. The postprocessor delivers the label and confidence to the visualizer. The visualizer starts to display the signal as soon as it gets access to the signal file, as shown in Figure 1 using the “Signal File” and “Visualizer” blocks. Once the visualizer receives the label and confidence for the latest epoch from the postprocessor, it overlays the decision and color codes that epoch. The visualizer uses red for seizure with the label SEIZ and green for the background class with the label BCKG. Once the streaming finishes, the system saves three files: a signal file in which the sample frames are saved in the order they were streamed, a time segmented event (TSE) file with the overall decisions and confidences, and a hypotheses (HYP) file that saves the label and confidence for each epoch. The user can plot the signal and decisions using the signal and HYP files with only the visualizer by enabling appropriate options. For comparing the performance of different stages of development, we used the test set of TUSZ v1.2.1 database. It contains 1015 EEG records of varying duration. The any-overlap performance [12] of the overall system shown in Figure 2 is 40.29% sensitivity with 5.77 FAs per 24 hours. For comparison, the previous state-of-the-art model developed on this database performed at 30.71% sensitivity with 6.77 FAs per 24 hours [3]. The individual performances of the deep learning phases are as follows: Phase 1’s (P1) performance is 39.46% sensitivity and 11.62 FAs per 24 hours, and Phase 2 detects seizures with 41.16% sensitivity and 11.69 FAs per 24 hours. We trained an LSTM model with the delayed features and the window-based normalization technique for developing the online system. Using the offline decoder and postprocessor, the model performed at 36.23% sensitivity with 9.52 FAs per 24 hours. The trained model was then evaluated with the online modules. The current performance of the overall online system is 45.80% sensitivity with 28.14 FAs per 24 hours. Table 2 summarizes the performances of these systems. The performance of the online system deviates from the offline P1 model because the online postprocessor fails to combine the events as the seizure probability fluctuates during an event. The modules in the online system add a total of 11.1 seconds of delay for processing each second of the data, as shown in Figure 3. In practice, we also count the time for loading the model and starting the visualizer block. When we consider these facts, the system consumes 15 seconds to display the first hypothesis. The system detects seizure onsets with an average latency of 15 seconds. Implementing an automatic seizure detection model in real time is not trivial. We used a variety of techniques such as the file locking mechanism, multithreading, circular buffers, real-time event decoding, and signal-decision plotting to realize the system. A video demonstrating the system is available at: https://www.isip.piconepress.com/projects/nsf_pfi_tt/resources/videos/realtime_eeg_analysis/v2.5.1/video_2.5.1.mp4. The final conference submission will include a more detailed analysis of the online performance of each module. ACKNOWLEDGMENTS Research reported in this publication was most recently supported by the National Science Foundation Partnership for Innovation award number IIP-1827565 and the Pennsylvania Commonwealth Universal Research Enhancement Program (PA CURE). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the official views of any of these organizations. REFERENCES [1] A. Craik, Y. He, and J. L. Contreras-Vidal, “Deep learning for electroencephalogram (EEG) classification tasks: a review,” J. Neural Eng., vol. 16, no. 3, p. 031001, 2019. https://doi.org/10.1088/1741-2552/ab0ab5. [2] A. C. Bridi, T. Q. Louro, and R. C. L. Da Silva, “Clinical Alarms in intensive care: implications of alarm fatigue for the safety of patients,” Rev. Lat. Am. Enfermagem, vol. 22, no. 6, p. 1034, 2014. https://doi.org/10.1590/0104-1169.3488.2513. [3] M. Golmohammadi, V. Shah, I. Obeid, and J. Picone, “Deep Learning Approaches for Automatic Seizure Detection from Scalp Electroencephalograms,” in Signal Processing in Medicine and Biology: Emerging Trends in Research and Applications, 1st ed., I. Obeid, I. Selesnick, and J. Picone, Eds. New York, New York, USA: Springer, 2020, pp. 233–274. https://doi.org/10.1007/978-3-030-36844-9_8. [4] “CFM Olympic Brainz Monitor.” [Online]. Available: https://newborncare.natus.com/products-services/newborn-care-products/newborn-brain-injury/cfm-olympic-brainz-monitor. [Accessed: 17-Jul-2020]. [5] M. L. Scheuer, S. B. Wilson, A. Antony, G. Ghearing, A. Urban, and A. I. Bagic, “Seizure Detection: Interreader Agreement and Detection Algorithm Assessments Using a Large Dataset,” J. Clin. Neurophysiol., 2020. https://doi.org/10.1097/WNP.0000000000000709. [6] A. Harati, M. Golmohammadi, S. Lopez, I. Obeid, and J. Picone, “Improved EEG Event Classification Using Differential Energy,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium, 2015, pp. 1–4. https://doi.org/10.1109/SPMB.2015.7405421. [7] V. Shah, C. Campbell, I. Obeid, and J. Picone, “Improved Spatio-Temporal Modeling in Automated Seizure Detection using Channel-Dependent Posteriors,” Neurocomputing, 2021. [8] W. Tatum, A. Husain, S. Benbadis, and P. Kaplan, Handbook of EEG Interpretation. New York City, New York, USA: Demos Medical Publishing, 2007. [9] D. P. Bovet and C. Marco, Understanding the Linux Kernel, 3rd ed. O’Reilly Media, Inc., 2005. https://www.oreilly.com/library/view/understanding-the-linux/0596005652/. [10] V. Shah et al., “The Temple University Hospital Seizure Detection Corpus,” Front. Neuroinform., vol. 12, pp. 1–6, 2018. https://doi.org/10.3389/fninf.2018.00083. [11] F. Pedregosa et al., “Scikit-learn: Machine Learning in Python,” J. Mach. Learn. Res., vol. 12, pp. 2825–2830, 2011. https://dl.acm.org/doi/10.5555/1953048.2078195. [12] J. Gotman, D. Flanagan, J. Zhang, and B. Rosenblatt, “Automatic seizure detection in the newborn: Methods and initial evaluation,” Electroencephalogr. Clin. Neurophysiol., vol. 103, no. 3, pp. 356–362, 1997. https://doi.org/10.1016/S0013-4694(97)00003-9. 
    more » « less