Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
17
datasets available to search
ShareScore release 0.9.0
Dataset results
17 results for “Linux kernel”
DongTing: A Large-scale Dataset for Anomaly Detection of the Linux Kernel
<p>DongTing is the first large-scale dataset dedicated to Linux kernel anomaly detection. The dataset covers Linux kernels released in the last five years and includes a total of 18,966 well-labeled normal and attack sequences. The entire dataset is 85 GB in size (after decompression). The attack data covers 26 major kernel releases and contains a total of 12,116 system call sequences collected from running 17,855 bug-triggering programs. The normal data comes from 6,850 normal programs in four kernel regression test suites. We maintain the dataset and source code in Zenodo and Github, respectively, and back up the dataset and code in Baidu netdisk.</p> <h3><strong>Dataset</strong></h3> <p>The dataset is stored at <a href="http://doi.org/10.5281/zenodo.6627050">http://doi.org/10.5281/zenodo.6627050</a></p> <ul> <li>The data includes `abnormal_data`, `normal_data`, `models`, `npz` and baseline data, with a total volume of nearly 87 GB (including 85 GB for abnormal data and normal data, it's after decompression files size).</li> <li>The `Abnormal_data` directory contains 12,116 files containing system call sequence for 26 kernel releases, and the `Normal_data` directory contains 6,850 files containing system call sequences collected from four regression test suites. All of which are raw sequences.</li> <li>CNN/RNN, LSTM, and Wavenet (three sets of hyperparameters per model) machine learning models are selected, the ECOD model (without hyperparameters) was also chosen for the evaluation of DT. DT_abnormal, DT_normal, ADFA-LD, and PLAID are used for training respectively. The results of DT training models are stored in the directory `Models-DongTing`, and the results of ADFA-LD and PLAID training models are stored in the directory `Models-Comparison`.</li> <li>The directory `npz `stores the encoded dataset of DongTing, ADFA-LD, and PLAID (sequence length varies from 8 to 4495), according to syscall_64.tbl in Linux kernel 5.17, including the training set, validation set, and test set.</li> <li>The file `Baseline.xlsx` contains all the information about DongTing dataset, which can be used in training machine learning models. For example, the whole dataset is randomly divided into three sets with the ratio of 80%:10%:10% (training: validation: test). The implementation of dataset division can be found in the source code.</li> </ul> <h3><strong>Source Code</strong></h3> <p><br>The source code for dataset development is stored at <a href="https://github.com/HNUSystemsLab/DongTing">https://github.com/HNUSystemsLab/DongTing</a> and the following is a brief introduction.</p> <ul> <li>The source code contains three folders, i.e., `Source Code Files`, `Documents` and `DB`, where `Documents `stores the detailed documents related to development, `DB` stores samples data, and `Source Code Files` stores the source code related to the development of our dataset.</li> <li>The detailed description about the source code can be found in `Documents/Documentation.pdf`. The document consists of four parts: environment requirements, database, program structure and working steps, model training and evaluation (including training and evaluation). It details the preparation of the environment, data import method, functional description of each file in the source code directory, how model training and evaluation work and other related contents.</li> </ul> <p>We additionally maintain the dataset and source code on Baidu.com <a href="https://pan.baidu.com/s/1vu1WGZpf2DqMIoyGayNu3w?pwd=dtds">https://pan.baidu.com/s/1vu1WGZpf2DqMIoyGayNu3w?pwd=dtds</a> to facilitate the access from China.</p> <p> </p> <h3>Tips: </h3> <p>If you find DongTing useful for your research, please cite the article as "DongTing: A large-scale dataset for anomaly detection of the Linux kernel".</p> <blockquote> <p><br>@article{DUAN2023111745,<br>title = {DongTing: A large-scale dataset for anomaly detection of the Linux kernel},<br>journal = {Journal of Systems and Software},<br>volume = {203},<br>pages = {111745},<br>year = {2023},<br>issn = {0164-1212},<br>doi = {https://doi.org/10.1016/j.jss.2023.111745},<br>url = {https://www.sciencedirect.com/science/article/pii/S0164121223001401},<br>author = {Guoyun Duan and Yuanzhi Fu and Minjie Cai and Hao Chen and Jianhua Sun}<br>}<br><br></p> </blockquote>
Linux Kernel 4.21 Call Graphs
<p>This is the<strong> Linux Kernel 4.21 Call Graphs</strong> created using <a href="http://github.com/dspinellis/cscout">CScout</a> containing the following graphs:</p> <ol> <li>File include graph (fgraph_I.txt) </li> <li>Compile Time Dependency Graph (fgraph_C.txt)</li> <li>Control Dependency Graph (through function calls) (fgraph_F_D.txt)</li> <li>Data Dependency Graph (through global variables) (fgraph_G.txt)</li> <li>Function and Macro Call Graph (cgraph.txt)</li> </ol> <p>Files are of the form</p> <p>foo.c boo.c</p> <p>which indicate a directed edge foo.c -> boo.c.</p> <p>The call graphs refer to <strong>all </strong>(ending with _all.txt) files or only the <strong>writable files.</strong> </p> <p>These graphs were produced by processing the Linux Kernel Codebase consisting of 20.3 million lines of source code. </p> <p>The results were produced on an Intel(R) Xeon(R) CPU E5-1410 0 @ 2.80GHz server with 64GB of RAM.</p> <p><strong>References: </strong></p> <p>1. Papachristou, Marios. "Software clusterings with vector semantics and the call graph." <em>Proceedings of the 2019 27th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering</em>. 2019.</p>
Linux kernels v2.6.14 v3.18.129 v6.14.2 processed by CScout
<p>This dataset contains three SQLite databases containing a full image of the <a href="https://www.kernel.org/">Linux kernel</a> in relational form, as generated by the <a href="https://www.spinellis.gr/cscout/">CScout</a> refactoring browser (version 03023d - 2024-08-30 and version afdd54 - 2025-04-19) piping its output to SQLite (version 3.46.0 2024-05-23.</p> <p>The datasets can be used to perform empirical research on the Linux kernel's source code, especially identifier semantics, which they represent in the form of equivalence classes. They form part of the replication package of the paper titled “You Are not Expected to Understand this: The Usage of the C Preprocessor in the Linux Kernel”.</p>
A dataset of Linux Kernel commits
<p>Dataset with metadata about more than 1,200,000 changes (commits) of the Linux kernel, corresponding to a period since 2005 to 2023, which<br>can be easily ingested in data analytics systems.<br><br>It also includes a list of more than 90,000 pairs of bug fixing changes and their corresponding bug introducing changes, labeled by developers of the Linux Kernel.</p>
A Multi-level Dataset of Linux Kernel Patchwork
<p>The dataset, source code, application and document of a MSR 2018 data showcase paper.</p> <p>Yulin Xu and Minghui Zhou. 2018. A Multi-level Dataset of Linux Kernel Patchwork. In MSR '18: MSR '18: 15th International Conference on Mining Software Repositories , May 28–29, 2018, Gothenburg, Sweden. ACM, New York, NY, USA, 4 pages. https://doi.org/10.1145/3196398.3196475</p> <p>Description of each files:</p> <p>The level-0 data is stored in the file `level-0.zip'.</p> <p>The level-1 and level-2 data are stored in the file `level-1 and level-2.7z'.</p> <p>The tool is the file `application.7z'.</p> <p>The source code of the tool is the file `source code.7z'.</p> <p>The document of the tool is the file `Usage of Application.pdf'.</p> <p>The document of the dataset is the file `Description of Tables.pdf'.</p>
A Dataset of Multiple Types of Linux Kernel Patches
<p>We share a dataset of a nine-year history of patches (666,550 patches produced from December 2008 to December 2017) and related discussion recorded by the LKML project on the Linux kernel patchwork. To better help future studies, we analyzed the review results of these patches, and marked three types: accepted patches, rejected patches, and patches that are rejected because of the communication problem.</p>
Data of the Paper "How to Communicate when Submitting Patches: an Empirical Study of the Linux Kernel"
<p>This repository includes the data of the paper "How to Communicate when Submitting Patches: an Empirical Study of the Linux Kernel".</p> <p>It contains three parts:</p> <ol> <li>Related online documents.</li> <li>Original questionnaire of the survey.</li> <li>Survey results.</li> </ol> <p> </p>
How Configurable is the Linux Kernel? Analyzing Two Decades of Feature-Model History
<p>Reproduction package for the TOSEM'25 paper "How Configurable is the Linux Kernel? Analyzing Two Decades of Feature-Model History"</p>
Reproduction package for the paper "Analyzing long-term maintenance releases of the Linux kernel"
<p>This is the reproduction package of the paper "Analyzing long-term maintenance releases of the Linux kernel". This package includes the datasets and the software for methods presented in the paper.</p>
Incorrect/Incomplete Security Bug Fixes in the Linux kernel
<p>Incorrect/Incomplete Security Bug Fixes in the Linux kernel</p>
the Linux kernel & Rust commits
<p>Historical commit data for the Linux kernel (2001/09/17-2023/11/22)/Rust (2010/06/23-2021/12/06)</p>
Software artifacts corresponding to the paper "Pragmatic Random Sampling of the Linux Kernel: Enhancing the Randomness and Correctness of the conf Tool"
<div><strong>Software artifacts corresponding to the paper "Pragmatic Random Sampling of the Linux Kernel: Enhancing the Randomness and Correctness of the conf Tool"</strong></div> <div> </div> <div> <div> <div>This repository is organized in two main folders:</div> <br> <div>1. <strong>randconfig+</strong> includes the source code of our tool randconfig+, which improves the randomness and correctness of the conf tool (a built-in Linux kernel tool for generating random samples of kernel configurations). It has two subfolders:</div> - <strong>source</strong>: includes the source code of randconfig+. <div> - <strong>bin</strong>: includes the compiled version of randconfig+.</div> <br> <div>2. <strong>experimental_validation</strong> includes the experimental validation of randconfig+ reported in the paper.</div> - <strong>systems</strong>: includes the 10 Linux Kernel versions analyzed in the paper. <div> - <strong>scripts_for_sample_generation</strong>: Bash shell scripts to generate the data.</div> <div> - <strong>data</strong>: includes the following data generated by the experimental validation:</div> <div> + Generated samples (subfolder configuration_samples). The samples are available in two formats: CSV and <a href="https://r4ds.hadley.nz/arrow">Apache Arrow</a>.</div> <div> + Entropies and number of distinct values per configuration option.</div> <div> + Data regarding the correctness of the samples and the time spent to generate them (subfolder correctness_and_runtime).</div> <div> - <strong>statistical_analysis</strong>: R scripts to perform the statistical analysis of the data (subfolder statistical_analysis).</div> <div> - <strong>plots</strong>: graphs produced from the data statistical analysis.</div> </div> </div>
Dataset of "How to Communicate when Submitting Patches: an Empirical Study of the Linux Kernel"
<p>This repository includes the data of the paper "How to Communicate when Submitting Patches: an Empirical Study of the Linux Kernel".</p> <p>It contains three parts:</p> <p>1) The related online documents.</p> <p>2) The survey results.</p> <p>3) The classified patches of LKML.</p> <p> </p> <p></p>
Dataset of "how to communicate when submitting patches: an Empirical Study of the Linux Kernel"
<p>This repository includes the data of the paper "How to Communicate when Submitting Patches: an Empirical Study of the Linux Kernel".</p> <p>It contains three parts:</p> <p>1) Related online documents.</p> <p>2) Original questionnaire of the survey.</p> <p>3) Survey results.</p>
Linux Kernel CTF trace
Trace obtained using lttng
Linux Kernel CTF trace
Trace obtained using lttng
Build and measurements of Linux kernel configurations across different versions
<p>With large scale and complex configurable systems, it is hard for users to choose the right combination of options (i.e., configurations) in order to obtain the wanted trade-off between functionality and performance goals such as speed or size. Machine learning can help in relating these goals to the configurable system options, and thus, predict the effect of options on the outcome, typically after a costly training step. However, many configurable systems evolve at such a rapid pace that it is impractical to retrain a new model from scratch for each new version. Taking the extreme case of the Linux kernel with its ≈ 14, 500 configuration options, we investigate how binary size predictions of kernel size degrade over successive versions (and how transfer learning can be adapted and applied to mitigate this degradation). </p> <p>We used and are sharing a unique and large dataset constituted of the binary sizes (compressed and non-compressed) of thousands of configurations for different versions of the kernel, spanning three years (4.13, 4.15, 4.20, 5.0, 5.4, 5.7, and 5.8). Overall, around 200K configurations over 10K+ options/features and 6 versions. </p> <p>This dataset has been used in the Transactions of Software Engineering (TSE) article "Transfer Learning Across Variants and Versions: The Case of Linux Kernel Size" (preprint: https://hal.inria.fr/hal-03358817) </p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.