Mapping the RISC-V Extensions Landscape (RISC-V Mentorship)

  • Internship
  • Remote

Mentor: Rafael Sene

Company: RISC-V International

The RISC-V Extensions Landscape is an open-source, interactive web tool that serves as a single searchable reference for the entire RISC-V Instruction Set Architecture ecosystem. It visualizes all ratified, draft, and discontinued extensions, displays per-instruction 32-bit encoding diagrams with match/mask values, and supports filtering by RISC-V profiles (RVA20, RVA22, RVA23, RVB23). The project is live at https://rpsene.github.io/riscv-extensions-landscape/ and built with React, Tailwind CSS, and Webpack.

The core challenge is that only 83 of 220 cataloged extensions (38%) have their instruction details fully mapped. The remaining 137 extensions exist in the catalog but lack instruction encodings, leaving significant gaps in supervisor-level extensions (trap, interrupt, memory), vector subsets, and vector cryptography bundles. Additionally, 104 extension tags from the upstream riscv-opcodes project remain unlinked due to naming mismatches between the source data and the catalog, preventing automated synchronization.

The mentee will work on three main tasks: (1) reconcile the 104 mismatched extension tags from riscv-opcodes with the catalog’s naming conventions, enabling the sync script to automatically populate instruction data; (2) research and map instructions for the remaining extensions, particularly supervisor-level extensions that define CSR behaviors and vector subset extensions that represent subsets of the V extension; and (3) improve the sync tooling to detect gaps and generate coverage reports automatically. The mentee will gain hands-on experience with the RISC-V ISA, open-source data engineering, and React-based visualization tools.

Repository URL: https://github.com/rpsene/riscv-extensions-landscape

Learning Objectives:

  • Introduction to open-source culture, workflows, and collaboration tools including Git, GitHub pull requests, code reviews, and issue tracking.
  • Understanding of the RISC-V Instruction Set Architecture, including base ISAs, standard and multi-letter extensions, instruction encodings (match/mask values), and the extension ratification lifecycle.
  • Proficiency in data engineering tasks such as reconciling datasets with different naming
    conventions, writing and maintaining synchronization scripts, and generating automated coverage reports.
  • Hands-on experience with a modern web development stack (React, Tailwind CSS, Webpack) by contributing to an interactive visualization tool used by the RISC-V community.
  • Ability to read and interpret ISA specification documents and translate them into structured, machine-readable data formats (JSON/JavaScript).
  • Experience with quality assurance practices including data validation, gap analysis, and encoding conflict detection across a large instruction set.
  • Regular one-on-one mentorship sessions covering code review feedback, technical guidance on RISC-V concepts, and career development discussions within the open-source hardware ecosystem.
  • Exposure to the RISC-V International community through participation in relevant technical group meetings and mailing lists.

Coding Challenge:

https://docs.google.com/document/d/18JR8IyC14-qh0o81m39XZ1VTnVbU71Z7nT-kXsQZDYw/edit?usp=sharing

To apply for this job please visit mentorship.lfx.linuxfoundation.org.