A binary-level solution for high-frequency code re-randomization.
A binary-level solution against advanced code-reuse attacks.
Secure and efficient Multivariant execution for binaries.
A practical context-sensitive CFI solution for binaries.
A binary-level solution against stack-based memory errors.
Binary and Malware Analysis
Disassembly analysis on full-Scale x86/x64 binaries.
Compiler-agnostic function detection for binaries.
Rage against the Machine Clear: A systematic analysis of Machine Clears and their implications for transient execution attacks.
Many-sided Rowhammer to bypass TRR mitigations on DDR4 DRAM chips.
Cache side-channel attacks over the network.
Rowhammer attacks on ECC-enabled systems.
Rowhammer attacks over the network and defenses.
Accelerating microarchitectural attacks with the GPU.
Cross-VM attacks abusing hardware vulnerabilities.
Deterministic Rowhammer exploitation on mobile devices.
How Google killed two-factor authentication.
Hacking Blind in the Spectre era.
Speculative data leaks across CPU cores are real.
RIDL is a new class of speculative execution attacks where an attacker can steal any “in-flight” data.
Employing the TLB in a novel sidechannel that doesn’t use the cache.
XLATE (translate) attacks reprogram the MMU to mount an indirect cache attack.
Memory deduplication as an advanced exploitation vector.
Thread spraying, allocation oracles, and defenses (MemSentry).
Side channeling the MMU for breaking ASLR in the browser.
Protecting memory deduplication against side-channel and Rowhammer attacks.
Run-time gadget-discovery framework.
Return-Oriented Programming without information disclosure.
Operating System with Integrated Recovery preventing Inconsistent State.
Software Testing and Sanitizers
Application-aware evolutionary fuzzing.
Kernel Multi-Variant eXecution.
Fast buffer overflow detection without branches.
Scalable use-after-free detection.
Practical mitigation of uninitialized read vulnerabilities.
Practical type confusion detection.
Validity of Research
A not-entirely-comprehensive of things you should not do in security research.
Benchmarking crimes in systems security research.
Prudent practices for designing malware experiments.