The specifications of mainstream processor architectures, such as Arm, x86, and RISC-V, underlie modern computing, as the targets of compilers, operating systems, and hypervisors. However, despite extensive research and tooling for instruction-set architecture (ISA) and relaxed-memory semantics, recently including systems features, there still do not exist integrated mathematical models that suffice for foundational formal verification, of concurrent architecture properties or of systems software. Previous proof-assistant work has had to substantially simplify the ISA semantics, the concurrency model, or both. We present ArchSem, an architecture-generic framework for architecture semantics, modularly combining ISA and concurrency models along a tractable interface of instruction-semantics effects, that covers a range of systems aspects. To do so, one has to handle many issues that were previously unclear, about the architectures themselves, the interface, the proper definition of reusable models, and the Rocq and Isabelle idioms required to make it usable. We instantiate it to the Arm-A and RISC-V instruction-set architectures and multiple concurrency models. We demonstrate usability for proof, despite the scale, by establishing that the Arm architecture (in a particular configuration) provides a provable virtual memory abstraction, with a combination of Rocq, Isabelle, and paper proof. Previous work provides further confirmation of usability: the AxSL program logic for Arm relaxed concurrency was proved sound above an earlier version of ArchSem. This establishes a basis for future proofs of architecture properties and systems software, above production architecture specifications.

Memory safety concerns continue to be a major source of security vulnerabilities. The CHERI architecture, as instantiated in prototype CHERI-RISC-V cores, the Arm Morello system, and Microsoft's CHERIoT embedded core, provides fine-grained memory access control through unforgeable hardware capabilities. The impact of CHERI on spatial memory safety is well understood. This paper systematically examines temporal memory safety within CHERI C -- a dialect of the C programming language for CHERI -- and proposes a formal approach to defining and ensuring it. In particular: 1) we examine the impact of five existing capability revocation mechanisms on CHERI C semantics and present a specialised object memory model tailored to CHERI C; 2) we introduce a new CHERI-specific pointer provenance tracking scheme; and 3) we formally define the security guarantees provided by this memory model, supported by a Coq proof of their correctness, expressed as invariants of the memory state.

To manage exceptions, software relies on a key architectural guarantee, precision: that exceptions appear to execute between instructions. However, this definition, dating back over 60 years, fundamentally assumes a sequential programmers model. Modern architectures such as Arm-A with programmer-observable relaxed behaviour make such a naive definition inadequate, and it is unclear exactly what guarantees programmers have on exception entry and exit.In this paper, we clarify the concepts needed to discuss exceptions in the relaxed-memory setting – a key aspect of precisely specifying the architectural interface between hardware and software. We explore the basic relaxed behaviour across exception boundaries, and the semantics of external aborts, using Arm-A as a representative modern architecture. We identify an important problem, present yet unexplored for decades: pinning down what it means for exceptions to be precise in a relaxed setting. We describe key phenomena that any definition should account for. We develop an axiomatic model for Arm-A precise exceptions, tooling for axiomatic model execution, and a library of tests. Finally we explore the relaxed semantics of software-generated interrupts, as used in sophisticated programming patterns, and sketch how they too could be modelled.

When designing new architectural security mechanisms, a key question is whether they actually provide the intended security, but this has historically been very hard to assess. One cannot gain much confidence by testing, as such mechanisms should provide protection in the presence of arbitrary unknown code. Previously, one also could not gain confidence by mechanised proof, as the scale of production instruction-set architecture (ISA) designs, many tens or hundreds of thousands of lines of specification, made that prohibitive. We focus in this paper especially on the secure encapsulation of software components, as supported by CHERI architectures in general and by the Arm Morello prototype architecture and hardware design in particular. Secure encapsulation is an essential security mechanism, for fault isolation and to constrain untrusted third-party code. It has previously often been implemented using virtual memory, but that does not scale to large numbers of compartments. Morello provides capability-based mechanisms that do scale, within a single address space. We prove a strong secure encapsulation property for an example of encapsulated code running on Morello, that holds in the presence of arbitrary untrusted code, above a full-scale sequential model of the Morello ISA. To do so, we build on, extend, and unify three orthogonal lines of previous work: the Cerise proof of such an encapsulation property for a highly idealised capability machine, expressed using a logical relation in Iris; the Islaris approach for reasoning about known code in production-scale ISAs; and the T-CHERI security properties of arbitrary Morello code, previously proved only for executions up to domain crossing. This demonstrates how one can prove such strong properties of security mechanisms for full-scale industry architectures.

Memory safety issues are a persistent source of security vulnerabilities, with conventional architectures and the C codebase chronically prone to exploitable errors. The CHERI research project has shown how one can provide radically improved security for that existing codebase with minimal modification, using unforgeable hardware capabilities in place of machine-word pointers in CHERI dialects of C, implemented as adaptions of Clang/LLVM and GCC. CHERI was first prototyped as extensions of MIPS and RISC-V; it is currently being evaluated by Arm and others with the Arm Morello experimental architecture, processor, and platform, to explore its potential for mass-market adoption, and by Microsoft in their CHERIoT design for embedded cores.

Abstract Architecture specifications such as Armv8-A and RISC-V are the ultimate foundation for software verification and the correctness criteria for hardware verification. They should define the allowed sequential and relaxed-memory concurrency behaviour of programs, but hitherto there has been no integration of full-scale instruction-set architecture (ISA) semantics with axiomatic concurrency models, either in mathematics or in tools. These ISA semantics can be surprisingly large and intricate, e.g. 100k $$+$$ + lines for Armv8-A. In this paper we present a tool, Isla, for computing the allowed behaviours of concurrent litmus tests with respect to full-scale ISA definitions, in the Sail language, and arbitrary axiomatic relaxed-memory concurrency models, in the Cat language. It is based on a generic symbolic engine for Sail ISA specifications. We equip the tool with a web interface to make it widely accessible, and illustrate and evaluate it for Armv8-A and RISC-V. The symbolic execution engine is valuable also for other verification tasks: it has been used in automated ISA test generation for the Arm Morello prototype architecture, extending Armv8-A with CHERI capabilities, and for Iris program-logic reasoning about binary code above the Armv8-A and RISC-V ISA specifications. By using full-scale and authoritative ISA semantics, Isla lets one evaluate litmus tests using arbitrary user instructions with high confidence. Moreover, because these ISA specifications give detailed and validated definitions of the sequential aspects of systems functionality, as used by hypervisors and operating systems, e.g. instruction fetch, exceptions, and address translation, our tool provides a basis for developing concurrency semantics for these. We demonstrate this for the Armv8-A instruction-fetch and virtual-memory models and examples of Simner et al.

Abstract Memory safety bugs continue to be a major source of security vulnerabilities in our critical infrastructure. The CHERI project has proposed extending conventional architectures with hardware-supported capabilities to enable fine-grained memory protection and scalable compartmentalisation, allowing historically memory-unsafe C and C++ to be adapted to deterministically mitigate large classes of vulnerabilities, while requiring only minor changes to existing system software sources. Arm is currently designing and building Morello, a CHERI-enabled prototype architecture, processor, SoC, and board, extending the high-performance Neoverse N1, to enable industrial evaluation of CHERI and pave the way for potential mass-market adoption. However, for such a major new security-oriented architecture feature, it is important to establish high confidence that it does provide the intended protections, and that cannot be done with conventional engineering techniques. In this paper we put the Morello architecture on a solid mathematical footing from the outset. We define the fundamental security property that Morello aims to provide, reachable capability monotonicity, and prove that the architecture definition satisfies it. This proof is mechanised in Isabelle/HOL, and applies to a translation of the official Arm specification of the Morello instruction-set architecture (ISA) into Isabelle. The main challenge is handling the complexity and scale of a production architecture: 62,000 lines of specification, translated to 210,000 lines of Isabelle. We do so by factoring the proof via a narrow abstraction capturing essential properties of arbitrary CHERI ISAs, expressed above a monadic intra-instruction semantics. We also develop a model-based test generator, which generates instruction-sequence tests that give good specification coverage, used in early testing of the Morello implementation and in Morello QEMU development, and we use Arm’s internal test suite to validate our model. This gives us machine-checked mathematical proofs of whole-ISA security properties of a full-scale industry architecture, at design-time. To the best of our knowledge, this is the first demonstration that that is feasible, and it significantly increases confidence in Morello.

Recent years have seen great advances towards verifying large-scale systems code. However, these verifications are usually based on hand-written assembly or machine-code semantics for the underlying architecture that only cover a small part of the instruction set architecture (ISA). In contrast, other recent work has used Sail to establish formal models for large real-world architectures, including Armv8-A and RISC-V, that are comprehensive (complete enough to boot an operating system or hypervisor) and authoritative (automatically derived from the Arm internal model and validated against the Arm validation suite, and adopted as the official formal specification by RISC-V International, respectively). But the scale and complexity of these models makes them challenging to use as a basis for verification.

Abstract Architecture specifications such as Armv8-A and RISC-V are the ultimate foundation for software verification and the correctness criteria for hardware verification. They should define the allowed sequential and relaxed-memory concurrency behaviour of programs, but hitherto there has been no integration of full-scale instruction-set architecture (ISA) semantics with axiomatic concurrency models, either in mathematics or in tools. These ISA semantics can be surprisingly large and intricate, e.g. 100k+ lines for Armv8-A. In this paper we present a tool, Isla, for computing the allowed behaviours of concurrent litmus tests with respect to full-scale ISA definitions, in Sail, and arbitrary axiomatic relaxed-memory concurrency models, in the Cat language. It is based on a generic symbolic engine for Sail ISA specifications, which should be valuable also for other verification tasks. We equip the tool with a web interface to make it widely accessible, and illustrate and evaluate it for Armv8-A and RISC-V. By using full-scale and authoritative ISA semantics, this lets one evaluate litmus tests using arbitrary user instructions with high confidence. Moreover, because these ISA specifications give detailed and validated definitions of the sequential aspects of systems functionality, as used by hypervisors and operating systems, e.g. instruction fetch, exceptions, and address translation, our tool provides a basis for developing concurrency semantics for these. We demonstrate this for the Armv8-A instruction-fetch model and self-modifying code examples of Simner et al.

The root causes of many security vulnerabilities include a pernicious combination of two problems, often regarded as inescapable aspects of computing. First, the protection mechanisms provided by the mainstream processor architecture and C/C++ language abstractions, dating back to the 1970s and before, provide only coarse-grain virtual-memory-based protection. Second, mainstream system engineering relies almost exclusively on test-and-debug methods, with (at best) prose specifications. These methods have historically sufficed commercially for much of the computer industry, but they fail to prevent large numbers of exploitable bugs, and the security problems that this causes are becoming ever more acute.In this paper we show how more rigorous engineering methods can be applied to the development of a new security-enhanced processor architecture, with its accompanying hardware implementation and software stack. We use formal models of the complete instruction-set architecture (ISA) at the heart of the design and engineering process, both in lightweight ways that support and improve normal engineering practice - as documentation, in emulators used as a test oracle for hardware and for running software, and for test generation - and for formal verification. We formalise key intended security properties of the design, and establish that these hold with mechanised proof. This is for the same complete ISA models (complete enough to boot operating systems), without idealisation.We do this for CHERI, an architecture with hardware capabilities that supports fine-grained memory protection and scalable secure compartmentalisation, while offering a smooth adoption path for existing software. CHERI is a maturing research architecture, developed since 2010, with work now underway on an Arm industrial prototype to explore its possible adoption in mass-market commercial processors. The rigorous engineering work described here has been an integral part of its development to date, enabling more rapid and confident experimentation, and boosting confidence in the design.

Dynamic binary translation (DBT) requires the implementation of load-link/store-conditional (LL/SC) primitives for guest systems that rely on this form of synchronization. When targeting, e.g., ×86 host systems, LL/SC guest instructions are typically emulated using atomic compare-and-swap (CAS) instructions on the host. Whilst this direct mapping is efficient, this approach is problematic due to subtle differences between LL/SC and CAS semantics. In this article, we demonstrate that this is a real problem, and we provide code examples that fail to execute correctly on QEMU and a commercial DBT system, which both use the CAS approach to LL/SC emulation. We then develop two novel and provably correct LL/SC emulation schemes: 1) a purely software-based scheme, which uses the DBT system's page translation cache for correctly selecting between fast, but unsynchronized, and slow, but fully synchronized memory accesses and 2) a hardware-accelerated scheme that leverages hardware transactional memory (HTM) provided by the host. We have implemented these two schemes in the Synopsys DesignWare ARC nSIM DBT system, and we evaluate our implementations against full applications, and targeted microbenchmarks. We demonstrate that our novel schemes are not only correct but also deliver competitive performance on-par or better than the widely used, but broken CAS scheme.

Architecture specifications notionally define the fundamental interface between hardware and software: the envelope of allowed behaviour for processor implementations, and the basic assumptions for software development and verification. But in practice, they are typically prose and pseudocode documents, not rigorous or executable artifacts, leaving software and verification on shaky ground. In this paper, we present rigorous semantic models for the sequential behaviour of large parts of the mainstream ARMv8-A, RISC-V, and MIPS architectures, and the research CHERI-MIPS architecture, that are complete enough to boot operating systems, variously Linux, FreeBSD, or seL4. Our ARMv8-A models are automatically translated from authoritative ARM-internal definitions, and (in one variant) tested against the ARM Architecture Validation Suite. We do this using a custom language for ISA semantics, Sail, with a lightweight dependent type system, that supports automatic generation of emulator code in C and OCaml, and automatic generation of proof-assistant definitions for Isabelle, HOL4, and (currently only for MIPS) Coq. We use the former for validation, and to assess specification coverage. To demonstrate the usability of the latter, we prove (in Isabelle) correctness of a purely functional characterisation of ARMv8-A address translation. We moreover integrate the RISC-V model into the RMEM tool for (user-mode) relaxed-memory concurrency exploration. We prove (on paper) the soundness of the core Sail type system. We thereby take a big step towards making the architectural abstraction actually well-defined, establishing foundations for verification and reasoning.

Presenting large formal instruction set models as executable functions makes them accessible to engineers and useful for less formal purposes such as simulation. However, it is more difficult to extract information about the behaviour of individual instructions for reasoning. We present a method which combines symbolic evaluation and symbolic execution techniques to provide a rule-based view of instruction behaviour, with particular application to automatic test generation for large MIPS-like models.