Analyzing Bytes: Pre-Disassembly Static Binary Analysis
summary
Built a pre-disassembly analyzer (AMaze) that runs a backward dataflow on an super-set control-flow graph that treats every byte offset as a potential instruction, computes path properties like stack-depth and register-preservation, and returns results for every offset in O(n) per binary. Evaluation on standard corpora shows ~700–1000× speedups (300k offsets in ~1 minute, ~2 GB RAM) and accuracy comparable to Angr/Ghidra/Dyninst by exploiting backward path fidelities.
AI-written summary · about
chapters (10)
- 0:00Binary analysis challenges
- 1:05Reverse engineering limits
- 2:51Disassembly scalability issues
- 4:33Code-data separation struggles
- 6:16Path properties for binary analysis
- 8:43Static analysis challenges
- 9:46Backward analysis linear time
- 11:55Backward path fidelities
- 13:43Evaluation methodology
- 15:54Conclusion and demo
related talks
21mDemystify Today's Binary Disassembling and How Modern ABI Makes it Easier
23mAn In-Depth Analysis of Disassembly on Full-Scale x86/x64 Binaries
45m"Advanced Malware Threat: handling obfuscated code using..."
38mStatic Code Analysis - A Behind-the-scenes Look
10mBacktraces in the mirror: mad science by secrets of elves & dwarves!!
01h 00mJust enough Assembly for Compiler Explorer
40mDynamic Binary Instrumentation Techniques to Address Native Code Obfuscation
48mHow to Build A Decompiler