Contents
SpecificationOASIS

OASIS VIRTIO

Virtual I/O Device (VIRTIO) Specification

VIRTIO defines standard virtual devices so an operating system inside a virtual machine can use services such as networking and storage through compatible drivers.

What it is

VIRTIO is an OASIS specification for virtual-device interfaces. A guest operating system runs a driver, while a backend supplies the corresponding device behaviour. This gives guests a common interface across different virtualisation implementations.

On an automotive high-performance computer (HPC), this can support an operating system accessing virtual networking or storage without directly controlling the physical peripheral. The backend may run in a privileged service partition or another implementation chosen by the platform.

Who it applies to

VIRTIO primarily concerns teams developing or integrating guest drivers, hypervisors and device backends. Platform architects also need it when allocating devices and services among virtual machines.

It is a technical specification, with adoption determined by the platform design or contractual requirements. Using virtualisation does not automatically require VIRTIO. A project must identify which interfaces it actually uses and who owns compatibility on each side.

Application developers usually encounter it indirectly through operating-system services. Integration engineers need enough understanding to distinguish application failures from driver, backend or resource-allocation problems.

What it covers

The specification defines common device mechanisms, transport-specific interfaces and requirements for individual device types. Its building blocks include:

Mechanism Purpose
Feature negotiation Establish which optional capabilities the driver and device will use
Virtqueues Exchange buffer descriptions and completion information through shared memory
Status and reset Coordinate initialisation and device lifecycle
Transport bindings Define discovery and access through mechanisms such as PCI or memory-mapped I/O
Device-specific interfaces Define behaviour for device classes such as network and block devices

Split and packed virtqueues provide different queue layouts. Supporting VIRTIO does not imply supporting every device type, optional feature or transport.

How it fits with other documents

VIRTIO describes the interface visible to the driver. The selected hypervisor, operating system and backend documentation must explain which parts their implementations support and how to configure them.

For an automotive project, apply this alongside the platform architecture, resource budget and safety and cybersecurity requirements. Those project documents must address scheduling, isolation, access permissions and failure containment. Interface compatibility does not by itself demonstrate that a shared device service meets vehicle-level requirements.

In practice

Start with a compatibility matrix covering the guest driver, backend, specification revision, transport, queue layout and enabled features. Record the negotiated configuration during integration so failures can be reproduced after software changes.

For example, when a cockpit guest uses virtual storage, trace the complete path from its driver through the backend to the physical storage controller. Identify which other guests share that path, who can reset it and what happens when the backend becomes unavailable.

Test startup, reset, backend failure, malformed requests and operation under contention from other guests. Measure latency and throughput with representative workloads. Review shared-memory access and buffer lifetime carefully: an interface that works during normal operation can still fail during restart or overload.

Make ownership explicit. Assign responsibility for driver updates, backend fixes, compatibility testing and failure diagnosis before the platform enters maintenance.

What it means for your discipline

  • HPC & application softwareCore

    Guest-driver and device-backend developers implement the interface and its required behaviour. Daily work includes feature negotiation, queue handling, memory ordering and recovery.

  • E/E architectureCore

    Platform architects select which virtual devices each guest receives and where their backends run. They define supported transports, features and dependencies between partitions.

  • CybersecurityPartial

    Security engineers assess the trust boundary between guests and backends. They review buffer validation, shared-memory permissions and how a compromised guest could affect other workloads.

  • Functional safetyPartial

    Safety engineers assess whether shared backends and resources can interfere with safety-related workloads. VIRTIO compatibility alone does not establish timing guarantees or freedom from interference.

  • Hardware & siliconPartial

    Hardware engineers help establish the memory, interrupt and DMA arrangements supporting virtual devices. They check that the processor and platform integration support the intended isolation and performance.

  • Networks & communicationPartial

    Network engineers must understand how virtual network interfaces connect guests to switching and physical interfaces. They assess traffic separation, buffering and contention across that path.

No direct stake: Base software · Diagnostics & lifecycle · Software update & OTA