1st Workshop on eBPF and Kernel Extensions

Call for Papers

eBPF (extended Berkeley Packet Filter) is an innovative technology that has been gaining popularity in the networking and operating system community for its flexibility, safety, and efficiency in programming end-host network and OS stacks. Despite the numerous advantages of eBPF, there are several research challenges in leveraging it for novel use cases. These challenges include the difficulty in integrating eBPF into current systems, the potential performance overhead when executed in the kernel, limitations of existing programming hooks and APIs, and the programming restrictions and challenges imposed by the need for safety as enforced by the eBPF verifier. To tackle these issues, a cross-disciplinary approach is necessary, combining techniques across network protocol design, programming languages, operating systems, compilers, hardware architecture, and formal verification.

The workshop aims to bring together experts and practitioners in the field of eBPF, end-host networking, and operating systems to discuss and present the latest advances to support and apply this cutting edge technology. The workshop will consist of invited talks, selected presentations, and a panel discussion to stimulate dialogue and exchange ideas about the use of eBPF to drive the next-generation of data center and cloud technologies. To this end, presenters will be selected from both industry and academia.

Topics of Interest

We welcome submissions including, but not limited to the following topics:

eBPF program development and deployment
eBPF in network performance analysis and optimization
eBPF in security and threat management
Previously unexplored, under-explored, and emerging use cases of eBPF
Techniques to optimize eBPF programs, for example static or dynamic optimizations
eBPF-based approaches for improving application efficiency
Simplifying eBPF programming
End-to-end network communication optimization using eBPF
Interaction between eBPF and end-host networking
eBPF use cases in high-performance networking
eBPF-based approaches to CPU scheduling
Security implications and innovations for end-host programmability using eBPF
Usage of hardware security or performance features to benefit eBPF
Approaches to improving the in-kernel JIT compiler for eBPF
Quantitative measurements or comparisons of eBPF to other programmable data plane technologies
Limitations of the eBPF verifier and approaches to overcome them
Approaches to scaling eBPF programs in large-scale distributed systems
Case studies and real-world deployments of eBPF and end-host networking
Future directions and challenges in eBPF and end-host networking
Techniques for software or hardware fault isolation in eBPF or related systems

Submission Instructions

The 1st workshop on eBPF and kernel extensions solicits submissions. We are looking for two-types of submissions:

Research papers of up to 6 pages, including all figures, tables and appendices. Submissions must be original, unpublished work, that have not been concurrently submitted to other venues. Workshop papers will appear in ACM DL as a part of the official proceedings, and are thus considered published work.
Extended abstracts, which are 2 pages in length (excluded references) submitted in the same format as the workshop papers.

All two types of papers can use as many additional pages as necessary for citations, and should be written using the two-column 10pt ACM sigconf format (https://github.com/scyue/latex-sigcomm18). All submissions are double-blind. The program committee will review papers to determine relevance to the workshop, quality, and on the likelihood that it will elicit discussion among the attendees. At least one author from each accepted submission must attend the workshop to present and discuss their work.

Please submit your paper via https://ebpf23.hotcrp.com/.

If you have any questions or problems with your submission, please get in touch with Sebastiano Miano (s.miano@qmul.ac.uk) or Srinivas Narayana Ganapathy (sn624@cs.rutgers.edu).