University of Houston

OpenACC is an emerging directive-based programming model for programming accelerators that typically enables non-expert programmers to achieve portable and productive performance of their applications. The model is gaining popularity and being used for accelerating many types of applications, ranging from molecular dynamics codes to particle physics models. It is critical to evaluate the correctness of the OpenACC compiler implementations and determine its conformance to the specification.

In this thesis, we present a scalable testing infrastructure that serves this purpose. We worked very closely with three main vendors that offer compiler support for OpenACC 2.0 and assisted them in identifying and resolving compiler bugs helping them improve the quality of their compilers. The testsuite aims to identify and resolve ambiguities within the OpenACC 2.0 specification. The framework of this testsuite is also robust enough to create test cases for the future releases.

We have implemented the testsuite consists of test cases for the most recent OpenACC 2.0 programming model, both for C and FORTRAN languages. We compare the results for several versions of each vendor compiler and demonstrate examples on identified bugs. Based on the interesting observations, we also give suggestions on improving the future specification./p>