Inclusion-Based Multi-level Pointer Analysis
Yingxia Cui, LI Long-shu, Sheng Yao
Abstract
Yingxia Cui, LI Long-shu, Sheng Yao
Abstract
A novel approach for points-to information analysis is presented, which can deal with not only the single dereferenced pointers, but also the multiple dereferenced pointers, arrays and pointer arithmetic. In order to analyze them, we extend to the language of inclusion constraints, and further give an inference system, which can be formalized as deriving all possible constraints, and there are no indirect constraints any more. Furthermore, an algorithm is presented, which comprises the single dereferenced pointers, multiple dereferenced pointers, arrays, pointer arithmetic analysis. Another advantage is to offload the work to a static analysis completely, which is a breakthrough of inclusion-based pointer analysis. So we are able to use two linear time algorithms to simplify constraint graph before analyzing, which is very crucial in the performance of the algorithm.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A novel approach for points-to information analysis is presented, which can deal with not only the single dereferenced pointers, but also the multiple dereferenced pointers, arrays and pointer arithmetic. In order to analyze them, we extend to the language of inclusion constraints, and further give an inference system, which can be formalized as deriving all possible constraints, and there are no indirect constraints any more. Furthermore, an algorithm is presented, which comprises the single dereferenced pointers, multiple dereferenced pointers, arrays, pointer arithmetic analysis. Another advantage is to offload the work to a static analysis completely, which is a breakthrough of inclusion-based pointer analysis. So we are able to use two linear time algorithms to simplify constraint graph before analyzing, which is very crucial in the performance of the algorithm.
Key concepts: Pointer analysis, Pointer (user interface), Computer science, Abstract interpretation, Programming language, Theoretical computer science, Static analysis, Algorithm