Section: New Results

WCET estimation

Participants : Damien Hardy, Hanbing Li, Isabelle Puaut, Erven Rohou.

Predicting the amount of resources required by embedded software is of prime importance for verifying that the system will fulfill its real-time and resource constraints. A particularly important point in hard real-time embedded systems is to predict the Worst-Case Execution Times (WCETs) of tasks, so that it can be proven that tasks temporal constraints (typically, deadlines) will be met. Our research concerns methods for obtaining automatically upper bounds of the execution times of applications on a given hardware. Our new results this year are on (i) multi-core architectures (ii) WCET estimation for faulty architectures (iii) traceability of flow information in compilers for WCET estimation.

WCET estimation and its interactions with compilation

On the comparison of deterministic and probabilistic WCET estimation techniques

Participants : Damien Hardy, Isabelle Puaut.

This is joint work with Jaume Abella, Eduardo Quinones and Francisco J. Cazorla from Barcelona Supercomputing Center

Several timing analysis techniques have been proposed to obtain Worst-Case Execution Time (WCET) estimates of applications running on a particular hardware. They can be classified into two classes of approaches: deterministic timing analysis techniques (DTA), that produce a unique WCET estimate, and probabilistic timing analysis techniques (PTA) that produce multiple WCET estimates with associated probabilities. Both approaches have their static (SDTA, SPTA) and measurement-based (MBDTA, MBPTA) variants. The lack of comparison figures among those techniques makes complex the selection of the most appropriate one.

This work [19] makes a first attempt towards comparing comprehensively SDTA, SPTA and MBPTA qualitatively and quantitatively, under different cache configurations implementing LRU and random replacement. We identify strengths and limitations of each technique depending on the characteristics of the program under analysis and the hardware platform, thus providing users with guidance on which approach to choose depending on their target application and hardware platform.

WCET estimation for architectures with faulty caches

Participants : Damien Hardy, Isabelle Puaut.

Technology scaling, used to increase performance, has the negative consequence of providing less reliable silicon primitives, resulting in an increase of the probability of failure of circuits, in particular for SRAM cells. While space redundancy techniques exist to recover from failures and provide fault-free chips, they will not be affordable anymore in the future due to their growing cost. Consequently, other approaches like fine grain disabling and reconfiguration of hardware elements (e.g. individual functional units or cache blocks) will become economically necessary. This fine-grain disabling will lead to degraded performance compared to a fault-free execution.

A common implicit assumption in all static worst-case execution time (WCET) estimation methods is that the target processor is not subject to faults. Their result is not safe anymore when using fine grain disabling of hardware components, which degrades performance.

In [16] a method that statically calculates a probabilistic WCET bound in the presence of permanent faults in instruction caches is provided. The method, from a given program, cache configuration and probability of cell failure, derives a probabilistic WCET bound. An essential benefit of our approach is that its probabilistic nature stems only from the probability associated with the presence of faults. By construction, the worst-case execution path cannot be missed, since it is determined using static analysis, extended to cope with permanent faults. This allows our method to be used in safety-critical real-time systems. The method is computationally tractable, since it avoids the exhaustive enumeration of all possible fault locations. Experimental results show that the proposed method accurately estimates WCETs in the presence of permanent faults compared to a method that explores all possible locations for faults. On the one hand, the proposed method allows to quantify the impact of permanent faults on WCET estimates for chips with a known probability of cell failure for the whole chip lifetime. On the other hand, and most importantly, our work can also be used in architectural exploration frameworks to select the most appropriate fault management mechanisms, for current and future chip designs.

Traceability of flow information for WCET estimation

Participants : Hanbing Li, Isabelle Puaut, Erven Rohou.

This research is part of the ANR W-SEPT project.

Control-flow information is mandatory for WCET estimation, to guarantee that programs terminate (e.g. provision of bounds for the number of loop iterations) but also to obtain tight estimates (e.g. identification of infeasible or mutually exclusive paths). Such flow information is expressed though annotations, that may be calculated automatically by program/model analysis, or provided manually.

The objective of this work is to address the challenging issue of the mapping and transformation of the flow information from high level down to machine code. In our recent work [21] , we have proposed a framework to systematically transform flow information from source code to machine code.

The framework defines a set of formulas to transform flow information for standard compiler optimizations. Transforming the flow information is done within the compiler, in parallel with transforming the code. There thus is no guessing what flow information have become, it is transformed along with the code. The framework is general enough to cover all linear flow constraints and all typical optimizations implemented in modern compilers. Our implementation in the LLVM compiler shows that we can improve the WCET of Malardalen benchmarks by 60% in average (up to 86%) by turning on optimizations. We also provide new insight on the impact of existing optimizations on the WCET.

Verified WCET estimation

Participant : Isabelle Puaut.

This is joint work with Andre Oliveira Maroneze, David Pichardie and Sandrine Blazy from the Celtique group at Inria/IRISA Rennes.

Current WCET estimation tools, even when based on sound static analysis techniques, are not verified. This may lead to bugs being accidentally introduced in the implementation. The main contribution of this work [13] , [26] is a formally verified WCET estimation tool operating over C code.

Our tool is integrated to the formally verified CompCert C compiler. It is composed of two main parts: a loop bound estimation and an Implicit Path Enumeration Technique (IPET)-based WCET calculation method. We evaluated the precision of the WCET estimates on a reference benchmark and obtained results which are competitive with state-of-the-art WCET estimation techniques. The code of our tool is automatically generated from its formal specification. Furthermore, machine-checked proofs ensure the estimated WCET is at least as large as the actual WCET.