Chapter 6
System Software Support for Hardware
Deficiency: Functions and Features
Abstract This chapter explains what is required to be changed or modified at the
level of system software (runtime system, languages) to address new properties of
computer system fault tolerance, resilience, and reliability. From the first principles,
we think through the system design attempting to reflect new system properties “as
early as possible”.
The rapidly developing area of IT applications in safety-critical real-time embedded
systems, up to now, was equipped by standard modifications of existing software
(Win CE [1], VxWorks [2], Linux, …) or for safety-critical applications designed
OS, such as Integrity OS [3].
Interestingly, no commercial fault-tolerant operating system is available on the
market. Integrity OS, for example, which is certified to be used in airborne applications (DO-178B, SKPP, EAL6+ High Robustness, according to their homepage
“the highest security level ever achieved for an operating system”), focuses as the
name implies on the integrity of the OS and its applications, i.e., memory protection, guaranteed resources (memory, CPU), and as they call it highest reliability by
guaranteeing system resources even in case of malfunctions.
The OS is responsible for isolating misbehaving applications due to malfunctions, but provides no further means of fault tolerance. The OS itself is thus prone to
malfunction and can according to our definition (GAFT) not be called fault tolerant
as it only provides application fault isolation.
None of these OS has been theoretically analyzed to answer the question what
system software can and must do to meet real-time constraints, while still providing
fault tolerance features of the system. This lack of theory is reflected in the misleading variety of solutions, huge overheads, and nearly impossible or prohibitively
expensive maintenance. Therefore, the aim of this research is
To design a new concept and its theoretical framework for further development of system
software for safety critical real time embedded systems, assuming possible hardware
deficiencies.
© Springer Nature Switzerland AG 2020
I. Schagaev et al., Software Design for Resilient Computer Systems,
https://doi.org/10.1007/978-3-030-21244-5_6
59
System Software Support for Hardware
Deficiency: Functions and Features
Abstract This chapter explains what is required to be changed or modified at the
level of system software (runtime system, languages) to address new properties of
computer system fault tolerance, resilience, and reliability. From the first principles,
we think through the system design attempting to reflect new system properties “as
early as possible”.
The rapidly developing area of IT applications in safety-critical real-time embedded
systems, up to now, was equipped by standard modifications of existing software
(Win CE [1], VxWorks [2], Linux, …) or for safety-critical applications designed
OS, such as Integrity OS [3].
Interestingly, no commercial fault-tolerant operating system is available on the
market. Integrity OS, for example, which is certified to be used in airborne applications (DO-178B, SKPP, EAL6+ High Robustness, according to their homepage
“the highest security level ever achieved for an operating system”), focuses as the
name implies on the integrity of the OS and its applications, i.e., memory protection, guaranteed resources (memory, CPU), and as they call it highest reliability by
guaranteeing system resources even in case of malfunctions.
The OS is responsible for isolating misbehaving applications due to malfunctions, but provides no further means of fault tolerance. The OS itself is thus prone to
malfunction and can according to our definition (GAFT) not be called fault tolerant
as it only provides application fault isolation.
None of these OS has been theoretically analyzed to answer the question what
system software can and must do to meet real-time constraints, while still providing
fault tolerance features of the system. This lack of theory is reflected in the misleading variety of solutions, huge overheads, and nearly impossible or prohibitively
expensive maintenance. Therefore, the aim of this research is
To design a new concept and its theoretical framework for further development of system
software for safety critical real time embedded systems, assuming possible hardware
deficiencies.
© Springer Nature Switzerland AG 2020
I. Schagaev et al., Software Design for Resilient Computer Systems,
https://doi.org/10.1007/978-3-030-21244-5_6
59