Dyn-Phi: A Trace-Informed Join Node for Triaging Opaque Predicates and Dead Branches in Binary Code

Opaque predicates inflate a binary's control-flow graph, and the symbolic execution that can prove them infeasible scales poorly, so analysts must decide where to spend it. Dyn-Phi is a triage formulation for that decision. At each use where several definitions of a location may arrive, it annotates the static reaching set with how often each was observed to arrive in execution traces. A never-observed definition becomes a cold candidate for a three-valued verifier, whose counter-examples are replayed as new evidence. We prove that observations lie inside the static set unless the CFG is incomplete, that dead definitions are cold, that code is removed only on the verifier's word, and that the loop terminates. An open implementation is evaluated on four corpora of 200 functions, two of them real gcc -O0 x86-64 code, with an exact verifier, paired bootstrap intervals, and two self-registered protocols. Most findings are negative. A cold-definition filter needs about one sixth of the exhaustive queries, but only because it skips cheap feasible ones; if proving infeasibility costs 100 times more, the saving is 9%. Ranking by the score is worse than address order, feedback adds 0.4 points of recall, and a trace-free path-sensitive static baseline removes a quarter of the dead pairs for free. The zero-hit bound fails under correlated traces. No obfuscator output or real symbolic verifier is evaluated. Code, tests, corpora and results: doi:10.5281/zenodo.23249925.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-10-09
DOI
https://doi.org/10.5281/zenodo.23199026
Primary Topic
Software Testing and Debugging Techniques
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

Dyn-Phi: A Trace-Informed Join Node for Triaging Opaque Predicates and Dead Branches in Binary Code

David Daryanto
Zenodo (CERN European Organization for Nuclear Research)
Software Testing and Debugging Techniques
preprint

Dyn-Phi: A Trace-Informed Join Node for Triaging Opaque Predicates and Dead Branches in Binary Code

David Daryanto
preprint en

Abstract

Opaque predicates inflate a binary's control-flow graph, and the symbolic execution that can prove them infeasible scales poorly, so analysts must decide where to spend it. Dyn-Phi is a triage formulation for that decision. At each use where several definitions of a location may arrive, it annotates the static reaching set with how often each was observed to arrive in execution traces. A never-observed definition becomes a cold candidate for a three-valued verifier, whose counter-examples are replayed as new evidence. We prove that observations lie inside the static set unless the CFG is incomplete, that dead definitions are cold, that code is removed only on the verifier's word, and that the loop terminates. An open implementation is evaluated on four corpora of 200 functions, two of them real gcc -O0 x86-64 code, with an exact verifier, paired bootstrap intervals, and two self-registered protocols. Most findings are negative. A cold-definition filter needs about one sixth of the exhaustive queries, but only because it skips cheap feasible ones; if proving infeasibility costs 100 times more, the saving is 9%. Ranking by the score is worse than address order, feedback adds 0.4 points of recall, and a trace-free path-sensitive static baseline removes a quarter of the dead pairs for free. The zero-hit bound fails under correlated traces. No obfuscator output or real symbolic verifier is evaluated. Code, tests, corpora and results: doi:10.5281/zenodo.23249925.

Zenodo (CERN European Organization for Nuclear Research)
Software Testing and Debugging Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.