Convite da Defesa do Projeto de Tese do Programa de Pós-Graduação em Ciência da Computação

A Coordenação do Programa de Pós-Graduação em Ciência da Computação tem a satisfação de convidá-lo para a Defesa de Projeto de Tese:


Scalable and Precise Dynamic Reconstruction of CFGs

Andrei Rimsa Álvares


The automatic recovery of a program’s high-level representation from its binary version is a well-studied problem. However, most of the solutions are based on purely static approaches: the bytes of an executable back are converted into control flow graphs (CFGs). This proposal shows that a dynamic analysis can be effective and useful as a standalone technique, and as a way to enhance the precision of static approaches. The experimental results provide evidence that completeness is achievable on many functions from industry-strong benchmarks. Experiments also indicate that it greatly enhances the ability of

DynInst to deal with code stripped of debugging information. These results were obtained with CFGGrind, a dynamic CFG reconstructor. CFGGrind recovers the complete CFG of 41% of Spec Cpu2017's procedures. When combined with CFGGrind, DynInst finds 36% more CFGs for cBench, and 21% more for Spec Cpu2017. Finally, CFGGrind is 7x faster than DCFG, and 1.30x faster than bfTrace.


Comissão Examinadora:

Prof. Fernando Magno Quintão Pereira - Orientador (DCC - UFMG)

Prof. Leonardo Barbosa e Oliveira (DCC - UFMG)

Prof. Marcos Augusto Menezes Vieira (DCC - UFMG)

Prof. Guido Costa Souza de Araújo (IC - UNICAMP)

Prof. José Nelson Amaral (DSC - University of Alberta)


19 de Novembro de 2019



Sala 2077 do ICEX