Compare less, defer more: scaling value-contexts based whole-program heap analyses
Citations
27 citations
Cites background from "Compare less, defer more: scaling v..."
...…2019, 2018; Jeong et al. 2017; Kastrinis and Smaragdakis 2013; Lhoták and Hendren 2006; Li et al. 2018b; Milanova et al. 2002, 2005; Smaragdakis and Balatsouras 2015; Smaragdakis et al. 2011, 2014; Sridharan et al. 2013; Tan et al. 2016, 2017; Thakur and Nandivada 2019; Thiessen and Lhoták 2017]....
[...]
13 citations
[...]
10 citations
Cites background from "Compare less, defer more: scaling v..."
...There have been prior works (not demand-driven) that present target client-specific solutions [31], improving the precision based on a complete set of queries [32]–[35]....
[...]
8 citations
Cites background or methods from "Compare less, defer more: scaling v..."
...In order to find whether a method or its callees satisfy Insight 1, we modify the multi-stage analysis approach (consisting of a pre-, a main- and a post-analysis) already in place for LSRV contexts [30], as discussed next....
[...]
...This is because of the identification of relevance and the notion of level-summarization, and the splitting of the overall approach into three stages: pre-analysis, main-analysis, and post-analysis [30]....
[...]
...However, as shown by Thakur and Nandivada [30],valcs does not scale for popular whole-program heap analyses....
[...]
...For a given analysis, Thakur and Nandivada [30] show that LSRV contexts (lsrv) only scale the corresponding valuecontexts based analysis, without affecting its precision....
[...]
...Thakur and Nandivada [30] identify the relevant portions of value contexts and summarize them to form the analysis-specific abstraction of LSRV contexts; their approach scales individual value-contexts based analyses, and hence call-strings based analyses, without compromising on the precision of the analysis....
[...]
5 citations
References
2,482 citations
1,561 citations
1,160 citations
618 citations