Polyspace R2021a
Polyspace R2021a: A Deep Dive into Features, Bug Fixes, and Workflow Enhancements
The release of Polyspace R2021a marked a significant milestone for engineers and embedded software developers striving for compliance with high-integrity standards (ISO 26262, DO-178C, IEC 61508). As part of the MathWorks Spring 2021 release, Polyspace R2021a didn’t just introduce incremental bug fixes; it redefined how static analysis integrates into CI/CD pipelines and agile development workflows.
This article explores the new features, performance improvements, and specific code-checking capabilities introduced in Polyspace R2021a, while providing guidance for teams looking to upgrade from legacy versions.
Integration with Simulink and Embedded Coder
Polyspace is not just for hand-written C/C++. In R2021a, the integration with Simulink and Embedded Coder matured significantly. polyspace r2021a
2. Polyspace Access Web Interface Enhancements
Polyspace Access—the web-based review and triage platform—received a redesigned review workflow:
- Batch review of similar findings across multiple files.
- Annotation export to Jira and GitLab issues.
- Role-based dashboards (developer vs. verification manager vs. certifier).
This bridges the gap between analysis results and traceability matrices required for DO-178C certification. Polyspace R2021a: A Deep Dive into Features, Bug
Polyspace Bug Finder (Static Analysis)
Bug Finder is a fast, scalable static analysis engine designed for developers during the coding phase. It identifies:
- Defects (e.g., null pointer dereferences, buffer overflows, division by zero)
- Security vulnerabilities (CWE, CERT C/C++/Java)
- Concurrency issues (data races, deadlocks)
- Code quality violations (complexity, unused variables)
In R2021a, Bug Finder improved its interprocedural analysis, enabling detection of defects across function boundaries without incurring the performance penalty of full formal verification. Integration with Simulink and Embedded Coder Polyspace is
Performance Benchmarks (R2021a vs. Prior Releases)
Based on internal MathWorks data and user reports (2021–2022):
| Metric | Polyspace R2020b | Polyspace R2021a | Improvement | |--------|------------------|------------------|--------------| | Code Prover analysis time (500k LOC, automotive) | 11.2 hrs | 8.1 hrs | 28% | | Memory usage (peak) | 32 GB | 24 GB | 25% | | Bug Finder false positive rate (MISRA C) | 12% | 8% | 33% | | Incremental reverification (single function change) | 100% of project | 28% of project | 72% reduction |
The incremental verification feature alone transformed Polyspace from a weekend-only tool to a commit-time gate.