Oliver Chang

12 papers A* 1A 1Journal 6Unranked 4
YearRankTypeTitle / Venue / Authors
2026 A* conf
AAAI
Oliver Chang
2025 J jnl
CoRR
Paschal C. Amusuo, Dongge Liu, Ricardo Andrés Calvo Méndez, Jonathan Metzman, Oliver Chang, James C. Davis
2025 J jnl
J. Imaging Inform. Medicine
Fatemeh Ghezloo, Oliver Chang, Stevan R. Knezevich, Kristin C. Shaw, Kia Gianni Thigpen, Lisa M. Reisch, Linda G. Shapiro, Joann G. Elmore
2024 J jnl
CoRR
Yuntong Zhang, Jiawei Wang, Dominic Berzin, Martin Mirchev, Dongge Liu, Abhishek Arya, Oliver Chang, Abhik Roychoudhury
2023 conf
FUZZING
Wentao Gao, Van-Thuan Pham, Dongge Liu, Oliver Chang, Toby Murray, Benjamin I. P. Rubinstein
2023 conf
SBFT
Dongge Liu, Jonathan Metzman, Marcel Böhme, Oliver Chang, Abhishek Arya
2023 J jnl
CoRR
Dongge Liu, Jonathan Metzman, Marcel Böhme, Oliver Chang, Abhishek Arya
2023 A conf
WACV
Kechun Liu, Beibin Li, Wenjun Wu, Caitlin J. May, Oliver Chang, Stevan Knezevich, Lisa M. Reisch, Joann G. Elmore, Linda G. Shapiro
2022 J jnl
J. Digit. Imaging
Shima Nofallah, Mojgan Mokhtari, Wenjun Wu, Sachin Mehta, Stevan Knezevich, Caitlin J. May, Oliver Chang, Annie C. Lee, Joann G. Elmore, Linda G. Shapiro
2021 conf
SSCI
Oliver Chang, Christiana Marchese, Jared Mejia, Anthony J. Clark
2021 conf
CVPR Workshops
Kechun Liu, Mojgan Mokhtari, Beibin Li, Shima Nofallah, Caitlin J. May, Oliver Chang, Stevan Knezevich, Joann G. Elmore, Linda G. Shapiro
2021 J jnl
IEEE Access
Wenjun Wu, Sachin Mehta, Shima Nofallah, Stevan Knezevich, Caitlin J. May, Oliver Chang, Joann G. Elmore, Linda G. Shapiro
redb/extractors/decompiler/bninja/analysis/medium_level.py
← Index redb/extractors/decompiler/bninja/analysis/medium_level.py python
import time

from binaryninja import (
    MediumLevelILOperation as MLIL_OP,
)

try:
    from ..function_type import FunctionTypeAnalysis
    from ..similarity.minhasher import MinHasher, TokenKind
    from ..utils.hashes import calculate_sha256, calculate_tlsh
    from .medium_level_normalization import MediumLevelNormalization
except ImportError:
    from redb.extractors.decompiler.bninja.analysis.medium_level_normalization import MediumLevelNormalization
    from redb.extractors.decompiler.bninja.similarity.minhasher import MinHasher
    from redb.extractors.decompiler.bninja.function_type import FunctionTypeAnalysis
    from redb.extractors.decompiler.bninja.utils.hashes import calculate_sha256, calculate_tlsh


_MLIL_CALL_OPS = (
    MLIL_OP.MLIL_CALL,
    MLIL_OP.MLIL_CALL_SSA,
    MLIL_OP.MLIL_CALL_UNTYPED,
    MLIL_OP.MLIL_CALL_UNTYPED_SSA,
    MLIL_OP.MLIL_TAILCALL,
    MLIL_OP.MLIL_TAILCALL_SSA,
    MLIL_OP.MLIL_TAILCALL_UNTYPED,
    MLIL_OP.MLIL_TAILCALL_UNTYPED_SSA,
)

_MLIL_CONTROL_FLOW_OPS = (
    MLIL_OP.MLIL_IF,
    MLIL_OP.MLIL_GOTO,
    MLIL_OP.MLIL_JUMP,
    MLIL_OP.MLIL_JUMP_TO,
    MLIL_OP.MLIL_RET,
    MLIL_OP.MLIL_RET_HINT,
    MLIL_OP.MLIL_NORET,
) + _MLIL_CALL_OPS


class MediumLevelAnalysis:
    def __init__(self, function, bv, logger):
        self.function = function
        self.name = function.name
        self.start = function.start
        self.mlil_func = function.mlil
        self.bv = bv
        self.logger = logger
        self.errors = []

    def log_error(self, message, function_name, address, exception=None, error_location="unknown"):
        error_msg = f"Error in function {function_name} at {address}: {message}"
        if exception:
            error_msg += f" - {str(exception)}"
        self.logger.error(error_msg)

        error = {
            "function_name": function_name,
            "function_address": str(address),
            "error_location": error_location,
            "error_message": message,
            "error_details": str(exception) if exception else "",
            "error_type": type(exception).__name__ if exception else "Unknown",
            "timestamp": int(time.time() * 1000),
        }
        self.errors.append(error)

    def _collect_mlil_skeleton_and_typed(self):
        mlil = self.mlil_func
        if not mlil:
            return [], [], [], []

        start = self.start
        norm = MediumLevelNormalization()

        skeleton = []
        skeleton_with_addr = []
        typed = []
        typed_with_addr = []

        for il in mlil.instructions:
            skel_norm = norm.normalize_instruction_all_levels(il)
            typed_norm = norm.normalize_instr_with_operands(il)

            skeleton.append(skel_norm)
            typed.append(typed_norm)

            offset = il.address - start
            if offset < 0:
                offset = 0

            skeleton_with_addr.append((offset, skel_norm))
            typed_with_addr.append((offset, typed_norm))

        return skeleton, skeleton_with_addr, typed, typed_with_addr

    def analyze(self):
        (
            instr_skeleton,
            body_mlil_skeleton_vector,
            instr_typed,
            body_mlil_typed_vector,
        ) = self._collect_mlil_skeleton_and_typed()

        instr_skeleton_str = str(instr_skeleton)
        sha256_skeleton = calculate_sha256(instr_skeleton_str)
        tlsh_skeleton = calculate_tlsh(instr_skeleton_str)

        instr_typed_str = str(instr_typed)
        sha256_typed = calculate_sha256(instr_typed_str)
        tlsh_typed = calculate_tlsh(instr_typed_str)

        seed = 0xdeadbeef
        minhash_mlil_skeleton = MinHasher(seed, self.mlil_func, TokenKind.MLIL).calculateMinHash()
        minhash_mlil_typed = MinHasher(seed, self.mlil_func, TokenKind.TYPED_MLIL).calculateMinHash()

        medium_level_json = {
            "function_address": self.start,
            "body_mlil_skeleton_vector": body_mlil_skeleton_vector,
            "sha256_mlil_skeleton": sha256_skeleton,
            "tlsh_mlil_skeleton": tlsh_skeleton,
            "minhash_mlil_skeleton": minhash_mlil_skeleton,
            "body_mlil_typed_vector": body_mlil_typed_vector,
            "sha256_mlil_typed": sha256_typed,
            "tlsh_mlil_typed": tlsh_typed,
            "minhash_mlil_typed": minhash_mlil_typed,
        }

        return medium_level_json, self.errors