Contents

Freedom from interference and partitioning

How software of different ASILs can share one processor without the lower-criticality part disturbing the higher one — in memory, timing and communication.

Why it exists

Developing all software on an ECU to ASIL D would be extremely expensive. Mixing ASIL D and QM software is only acceptable if the QM part provably cannot corrupt the ASIL part.

What kinds of interference exist

Kind Example Typical mechanism
Spatial (memory) QM code overwrites safety data MPU/MMU, SMMU, separate OS applications
Temporal (timing) QM task blocks the CPU Timing protection, budgets, separate cores
Communication Corrupted or delayed data between partitions E2E protection, checked interfaces

How it is implemented

On MCUs, the AUTOSAR OS separates software into OS applications with memory and timing protection. On HPCs, a hypervisor and the SoC’s SMMU and bandwidth controls isolate VMs; the safety island stays independent.

Best practices

  • Minimise shared resources; document every remaining one.
  • Verify isolation with fault injection tests.

Common pitfalls

  • Forgetting shared caches, interrupts or DMA as interference paths.

What it means for your discipline

  • Functional safety Core

    Defines the interference cases to rule out and argues sufficiency of the partitioning mechanisms.

  • Base software Core

    Configures OS applications, memory protection, timing protection and safe inter-partition communication.

  • HPC & application software Core

    On HPCs, hypervisors and OS mechanisms must provide the isolation between VMs and processes.

  • Hardware & silicon Core

    MPU/MMU, SMMU, bus firewalls and cache partitioning are the hardware basis.

  • Cybersecurity Partial

    The same isolation mechanisms also contain attacks — designs should serve both goals.

No direct stake: E/E architecture · Networks & communication · Diagnostics & lifecycle · Software update & OTA

Where it sits in the software stack

Generic layered software of a central HPC (application processor with hypervisor, plus a safety island) and of a classic MCU-based ECU (AUTOSAR Classic). Highlighted modules matter for this topic. Click any module to open its page.

HPC · application processor (MPU) and safety island (MCU)
  • Type-1 hypervisor — Isolates virtual machines through memory mappings, controlled device access and CPU scheduling budgets.
  • QNX / RT Linux — Separates safety-related processes and controls their memory access and execution time within the operating system.
  • Cortex-A cores · GPU/NPU · LPDDR — MMU, SMMU and shared-resource controls support isolation against memory corruption, DMA access and resource contention.
  • Own cores — Dedicated safety-island cores reduce interference from application-processor workloads; remaining shared resources still require analysis.
MCU-based ECU · AUTOSAR Classic
  • OS — OS applications, memory protection and timing protection prevent lower-criticality software from disturbing safety-related execution.
  • Microcontroller · HSM — The MPU and hardware access controls enforce memory boundaries; shared peripherals, DMA and interrupts require interference analysis.