本文へ移動
cccskills
無料GitHub で公開

device-driver-development

Use when developing, reviewing, debugging, porting, or engineering Linux kernel device drivers, kernel modules, device tree bindings, bus clients, subsystem drivers, DMA/IRQ/memory paths, userspace interfaces, or hardware bring-up plans.

インストール方法を見る

含まれるファイル(16)

  • SKILL.md4.2 KB
  • agent.md3.6 KB
  • agents/openai.yaml275 B
  • references/kernel-engineering-principles.md3.8 KB
  • references/knowledge-map.md2.9 KB
  • references/review-checklist.md2.7 KB
  • references/subsystem-matrix.md3.4 KB
  • references/tooling-matrix.md4.0 KB
  • scripts/driver_static_scan.py6.1 KB
  • scripts/module_smoke_build.sh848 B
  • scripts/tooling_probe.sh685 B
  • tasks/bringup-workflow.md2.6 KB
  • tasks/debugging-testing.md3.2 KB
  • tasks/implementation-patterns.md2.8 KB
  • tasks/step-verify-workflow.md2.7 KB
  • tasks/upstream-maintenance.md1.8 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Device Driver Development

Core Rule

Treat every driver task as hardware contract engineering: identify the kernel version, hardware bus, subsystem owner, execution context, lifetime model, and userspace ABI before writing code.

Knowledge Grounding

  • Use references/knowledge-map.md for the driver domain map and conceptual model.
  • Treat API details as version-sensitive; verify callback signatures, helpers, and bindings against the active kernel tree or current kernel docs.

First Pass

  1. Classify the device: char, platform, I2C, SPI, regmap, IIO, GPIO, IRQ controller, input, RTC, PWM, regulator, framebuffer, or NIC.
  2. Identify interface surfaces: /dev, sysfs, debugfs, ioctl, mmap, netdev, IIO buffers, input events, DT binding, or kernel consumer API.
  3. Choose probe/remove lifetime: prefer devm_*, explicit unwind only when ordering or shared state requires it.
  4. Decide concurrency boundaries: process context, atomic context, IRQ top half, threaded IRQ, workqueue, timer, DMA callback.
  5. Define evidence needed: datasheet registers, bus transaction examples, DT node, logs, scope/logic-analyzer traces, and test plan.

Required Reads By Task

  • New driver or hardware bring-up: tasks/bringup-workflow.md.
  • Implementing subsystem callbacks: tasks/implementation-patterns.md and references/subsystem-matrix.md.
  • Memory, MMIO, mmap, DMA, IRQ, locking: references/kernel-engineering-principles.md.
  • Debugging, validation, or intermittent behavior: tasks/debugging-testing.md.
  • Toolchain, QEMU, KGDB, bpftrace, Smatch, Coccinelle, Syzkaller, Bootlin: references/tooling-matrix.md.
  • Auditable engineering flow: tasks/step-verify-workflow.md.
  • Upstream-ready patches or long-term maintenance: tasks/upstream-maintenance.md.
  • Final review before handoff: references/review-checklist.md.

Design Heuristics

  • Prefer an existing kernel subsystem over a private char driver when the device matches a framework.
  • Keep policy in userspace and mechanism in kernel; avoid inventing ABI without necessity.
  • Never sleep in atomic context; never hold spinlocks across bus calls that may sleep.
  • Use copy_{to,from}_user() only at ABI boundaries and handle partial copy/error paths.
  • For MMIO use devm_ioremap_resource() plus accessor APIs; do not dereference __iomem.
  • For register devices on I2C/SPI, consider regmap before custom read/write helpers.
  • For DT, model hardware, not Linux driver internals; document compatibles and supplies/clocks/IRQs/GPIOs.

Script Helpers

  • Run scripts/driver_static_scan.py <path> for heuristic review of C, Kconfig, Makefile, DTS, and binding files.
  • Run scripts/module_smoke_build.sh <module-dir> [kernel-build-dir] for out-of-tree build smoke checks.
  • Run scripts/tooling_probe.sh to inspect local build, static-analysis, tracing, fuzzing, QEMU, and KGDB tool availability.

Verification Gate

  • Build with warnings enabled and, when available, sparse, Smatch, Coccinelle, checkpatch.pl, dt_binding_check, and subsystem selftests.
  • Test load/unload, probe/remove failure paths, hot/unplug or unbind where applicable, suspend/resume if resources are power-managed.
  • Capture runtime evidence: dmesg, /proc/interrupts, sysfs state, tracepoints/ftrace/bpftrace, bus transactions, and userspace API results.
  • Use QEMU/KGDB for isolated crash reproduction and Syzkaller for fuzzable ABI surfaces when practical.
  • Do not claim correctness from compilation alone; tie each claim to a code path, runtime observation, or documented subsystem contract.

Output Standard

When answering, name assumptions, selected subsystem, key APIs, file layout, validation commands, and residual hardware risks. If source evidence is missing, say what must be measured or read next.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Use when evaluating, designing, drafting, revising, stress-testing, or preparing academic articles, empirical research reports, literature reviews, systematic reviews, meta-analyses, journal submissions, reviewer responses, research methods, sampling, measurement, analysis, publication readiness, or citation/source audits.

日本語の概要は準備中です。原文の説明を表示しています。

oghie/skillsets92026年10月3日 更新

Android device offensive security, forensic extraction, and mobile app exploitation. Use this whenever the user needs to: break into or bypass the lock screen of an Android phone, extract data from a seized Android device for evidence, analyze or reverse-engineer an APK to find vulnerabilities, intercept communications on an Android device, exploit Android apps via deeplinks, steal crypto wallets or app data, set up rogue WiFi to target a phone, or plan any mobile device investigation or penetration test. Also trigger when the user mentions Android forensics, mobile pentesting, law enforcement phone extraction, device exploitation, WhatsApp or Telegram extraction from a phone, Android app reversing, CVE research for Android, or needing a chain of custody for mobile evidence. Even if the user describes the scenario casually or in another language, if they need to gain access to or pull data from an Android phone, use this skill.

日本語の概要は準備中です。原文の説明を表示しています。

oghie/skillsets92026年10月3日 更新

Use when performing cybersecurity operations - offensive or defensive - including threat detection and hunting, SOC operations, incident response, digital forensics and DFIR, malware analysis and reverse engineering, threat intelligence, cloud/container/network/application/API security, identity-access and zero-trust, OT/ICS and endpoint security, vulnerability and supply-chain management, DevSecOps, cryptography, AI security, compliance and governance, and red-team or penetration testing under signed rules of engagement; routes to one of 800+ technique playbooks mapped to MITRE ATT&CK, D3FEND, and NIST CSF.

日本語の概要は準備中です。原文の説明を表示しています。

oghie/skillsets92026年10月3日 更新

Use when designing, reviewing, modifying, migrating, optimizing, securing, operating, troubleshooting, or decommissioning data architectures across data model exploration, SQL, ISO SQL conformance, NoSQL, vector, graph, document, time-series, NewSQL, search, in-memory, SQLite, CDC/Debezium, DBMS storage/query/recovery internals, database incident prevention, hardware acceleration, cache, big-data, lakehouse, and warehouse systems.

日本語の概要は準備中です。原文の説明を表示しています。

oghie/skillsets92026年10月3日 更新

Use when designing, implementing, reviewing, debugging, testing, profiling, or porting concurrent, parallel, distributed, high-performance, network dataplane, kernel-bypass, or real-time software, including locks, semaphores, monitors, MPI, CSP rendezvous, DPDK, VPP, eBPF, XDP, AF_XDP, SR-IOV, RSS, RoCEv2/RDMA, RMS/EDF, priority inversion, and environment selection from silicon to edge.

日本語の概要は準備中です。原文の説明を表示しています。

oghie/skillsets92026年10月3日 更新

Use when designing, evaluating, documenting, modernizing, reviewing, refactoring, or implementing software architecture and code-level structure, including requirements, context, architecture style selection, microservices patterns, WebAssembly/WASI/Component Model, UMA portable capabilities, runtime contracts, capability graphs, governed MCP/agent orchestration, clean code, code smells, Rust crate/library API architecture, Cargo feature/MSRV/SemVer policy, unsafe/FFI/no_std/library testing, modularity, views, NFR tactics, IAM/auth, security patterns, cost estimation, architecture-as-code fitness functions, ADR/RFC work, and engineering execution plans.

日本語の概要は準備中です。原文の説明を表示しています。

oghie/skillsets92026年10月3日 更新

oghie のスキルをすべて見る

このスキルの問題を報告する