In the design cycle of a microprocessor core, the unit is usually refined through a series of subsequent steps. To deliver a flaw free unit at the end of the process, in each stage a verification step is required. While it would be useful to automatically develop the set of test programs for verification concurrently to the design, in most of the existing approach verification is performed manually and starting from scratch. This paper presented a methodology for the automatic completion and refinement of existing verification programs. It shows a new technique for allowing a Genetic Programming-based framework to import an existing test-program set and assimilate it for further test generation. A case study is considered, in which a sample pipelined processor is used, and new test programs are generated starting from existing functional ones. Different metrics are targeted, and preliminary results are reported, showing the effectiveness of the method with respect to a pure random approach.

Automatic completion and refinement of verification sets for microprocessor cores / Sanchez Sanchez, Edgar Ernesto; Sonza Reorda, Matteo; Squillero, Giovanni. - STAMPA. - 3449:(2005), pp. 205-214. ((Intervento presentato al convegno EvoWorkshops 2005: Applications of Evolutionary Computing [10.1007/978-3-540-32003-6_21].

Automatic completion and refinement of verification sets for microprocessor cores

SANCHEZ SANCHEZ, EDGAR ERNESTO;SONZA REORDA, Matteo;SQUILLERO, Giovanni
2005

Abstract

In the design cycle of a microprocessor core, the unit is usually refined through a series of subsequent steps. To deliver a flaw free unit at the end of the process, in each stage a verification step is required. While it would be useful to automatically develop the set of test programs for verification concurrently to the design, in most of the existing approach verification is performed manually and starting from scratch. This paper presented a methodology for the automatic completion and refinement of existing verification programs. It shows a new technique for allowing a Genetic Programming-based framework to import an existing test-program set and assimilate it for further test generation. A case study is considered, in which a sample pipelined processor is used, and new test programs are generated starting from existing functional ones. Different metrics are targeted, and preliminary results are reported, showing the effectiveness of the method with respect to a pure random approach.
978-3-540-25396-9
File in questo prodotto:
File Dimensione Formato  
Sánchez2005_Chapter_AutomaticCompletionAndRefineme.pdf

non disponibili

Tipologia: 2a Post-print versione editoriale / Version of Record
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 108.14 kB
Formato Adobe PDF
108.14 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

Caricamento pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11583/2501279