The use of standard libraries is a common practice in software development, highlighting the critical need to enhance library Application Binary Interface (ABI) compatibility across diverse Instruction Set Architectures (ISAs) in binary translation. While standard library functions’ Application Programming Interface (API) remains consistent across ISAs, their detailed implementations, particularly those involving composite data types, exhibit certain variations. These discrepancies result in ABI inconsistencies for identical function APIs across ISAs. Failure to resolve such ABI mismatches during binary translation can lead to misinterpretation of data passed via the ABI on the target ISA, ultimately causing application crashes. This paper introduces CDMF, a C ross-ISA D ata M orphing F ramework to enhance library ABI compatibility in static binary translation (SBT). CDMF automates the detection of cross-ISA library ABI incompatibilities, addressing challenges that previously demanded manual intervention. Our evaluation shows that integrating CDMF with EFACTMC improves the translation success rates by 33. 3% on SPEC CPU ® 2006 C benchmarks. Furthermore, we present EC-SBT, a retargetable SBT enhanced by CDMF, which efficiently migrates x86₆4 binaries to AArch64. Compared with Box64, EC-SBT is 1. 74× faster on EEMBC and is 1. 23× faster on SPEC CPU ® 2006 C benchmarks, showcasing its efficiency.
Zhang et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: