FX!32 a profile-directed binary translator

Chernoff, Anton; Herdeg, M.; Hookway, Ray; Reeve, C; Rubin, Norman; Tye, Tony; Yadavalli, S. Bharadwaj; Yates, J. K.

doi:10.1109/40.671403

Cited by 155 publications

(101 citation statements)

References 3 publications

Supporting

Mentioning

101

Contrasting

Order By: Relevance

“…In its general structure LIntel is similar to many other binary translation and optimization systems described before [11], [12], [13] and is very close to Transmeta's Code Morphing Software [14], [15]. As any other VM-level binary translation system, it has to solve the problem of efficient sharing of computational resources between translation and execution of original binary codes.…”

Section: Lintelmentioning

confidence: 99%

Background Optimization in Full System Binary Translation

Sokolov¹,

Ermolovich²

2011

Proceedings of the Spring/Summer Young Researchers' Colloquium on Software Engineering

View full text Add to dashboard Cite

Abstract-Binary translation and dynamic optimization are widely used to provide compatibility between legacy and promising upcoming architectures on the level of executable binary codes. Dynamic optimization is one of the key contributors to dynamic binary translation system performance. At the same time it can be a major source of overhead, both in terms of CPU cycles and whole system latency, as long as optimization time is included in the execution time of the application under translation. One of the solutions that allow to eliminate dynamic optimization overhead is to perform optimization simultaneously with the execution, in a separate thread. In the paper we present implementation of this technique in full system dynamic binary translator. For this purpose, an infrastructure for multithreaded execution was implemented in binary translation system. This allowed running dynamic optimization in a separate thread independently of and concurrently with the main thread of execution of binary codes under translation. Depending on the computational resources available, this is achieved whether by interleaving the two threads on a single processor core or by moving optimization thread to an underutilized processor core. In the first case the latency introduced to the system by a computational intensive dynamic optimization is reduced. In the second case overlapping of execution and optimization threads also results in elimination of optimization time from the total execution time of original binary codes.

show abstract

Section: Lintelmentioning

confidence: 99%

Background Optimization in Full System Binary Translation

Sokolov¹,

Ermolovich²

2011

Proceedings of the Spring/Summer Young Researchers' Colloquium on Software Engineering

View full text Add to dashboard Cite

show abstract

“…SIND can be built so that it allocates no dynamic memory. In this case, SIND occupies 440K of space 3 . This footprint will grow, however, as SIND is made self-contained (as explained in 7.2).…”

Section: Memory Footprint Of Sindmentioning

confidence: 99%

“…Many previous systems were created by companies as either an internal research tool [1] or as a commercial product [3], and consequently were never released. As the discussion in Chapter 2 reveals, many of these systems also have deficiencies or peculiar requirements.…”

Section: Why Another One?mentioning

confidence: 99%

“…FX 32 [3] is a dynamic translation program from DEC. It is designed to translate IA32 binaries to Alpha code at runtime.…”

Section: Fx 32mentioning

confidence: 99%

“…These solutions employ compiler extensions [20,24], hardware characteristics [25,32], kernel modifications [37,32], library modifications [3], or static analysis [41] to prevent and detect exploitation of buffer overflow vulnerabilities.…”

Section: Other Defenses Against Buffer Overflowsmentioning

confidence: 99%

See 2 more Smart Citations

Automated Diversity in Computer Systems

Forrest¹

2005

View full text Add to dashboard Cite

REPORT DOCUMENTATION PAGEForm Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and Attackers penetrate a large number of computers by exploiting common vulnerabilities. The objective of this effort is to address this internet-wide weakness by introducing diversity into computers so that a successful attack on one computer does not necessarily work on another one, even though it may be running identical software. Introducing diversity into computers---even those running identical software---so that successful attacks on one computer do not necessarily work on others. Diversity is one aspect of the kind of adaptive and robust methods used routinely in biological systems. Other adaptations beyond diversity were developed under this effort in the same spirit of increasing the resilience of our computing infrastructure. Summary of results:1. We explored three potential diversity mechanisms: Dynamic translation of machine code (to defend against code-injection), randomizing the system-call interface (to defend against codeinjection attacks), and evolving diverse implementations of TCP resource management policies (to defend against Denial of Service (DoS) attacks). Of these, we focused primarily on the first mechanism---dynamic translation. Dynamic translation of machine code is used to achieve Randomized Instruction Set Emulation (RISE), which thwarts malicious code injection attacks by making injected code appear random and thus illegal to the native processor.2. We developed a prototype implementation of machine code randomization, called RISE (randomized instruction set emulation). RISE is available under GPL licensing from: http://www.cs.unm.edu/~gbarrant. We tested RISE's performance at stopping attacks using the Core Impact testing software. We conducted extensive experiments, both to test the effectiveness of RISE at stopping attacks and to test the safety of executing random sequences of instructions (during a code-injection attack, the attack code is effectively randomized and we wanted to know how quickly and with how much certainty such randomized code would fail).3. We also continued the development of a dynamic translation tool (SIND) to facilitate dynamic translation diversity on RISC platforms. These results were reported in Trek Palmer's MS Thesis.4. We performed experimental validation and probabilistic analysis of RISE, for both RISC and CISC architectures, demonstrating and explaining the efficacy of the RISE approach. RISE succeeds in converting a very high p...

show abstract

x86‐Android performance improvement for x86 smart mobile devices

Choi

Lim

2014

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYWith more and more smartphones with Intel inside available, the compatibility issues of Android applications have been rising. The efficiency on the emulated architecture is best translated into efficiency on the target machine if target instructions also operated on register operands. However, conventional binary translators of popular binary translators do not take into account instruction dependency among two or more basic blocks. This results in performance degradation because of intertranslation block dependency. Actually, x86 binary translators do not take into account instruction dependency within a basic block. Binary translation makes use of one or two registers for the majority of translation blocks. This is because the translation block corresponds to a guest instruction, which in turn the amount of work is not large. Even though there are no dependencies between translation blocks, false dependencies are generated by the same register usage order. In order to maximize the parallelism within a basic block, we maintain two different register usage orders, applying them in turn. This approach resolves the problem and outperforms conventional approach by up to around 27% for some cases.

show abstract

FX!32 a profile-directed binary translator

Cited by 155 publications

References 3 publications

Background Optimization in Full System Binary Translation

Background Optimization in Full System Binary Translation

Automated Diversity in Computer Systems

x86‐Android performance improvement for x86 smart mobile devices

Contact Info

Product

Resources

About