Support for detecting some loaders (#113)

* Support for detecting some loaders Signed-off-by: Prabhu Subramanian <prabhu@appthreat.com> * Support for detecting some loaders Signed-off-by: Prabhu Subramanian <prabhu@appthreat.com> * Support for detecting some loaders Signed-off-by: Prabhu Subramanian <prabhu@appthreat.com> --------- Signed-off-by: Prabhu Subramanian <prabhu@appthreat.com>
owasp-dep-scan · Jul 29, 2024 · d574ea1 · d574ea1
1 parent 9759641
commit d574ea1
Show file tree

Hide file tree

Showing 14 changed files with 208 additions and 40 deletions.
diff --git a/Info.plist b/Info.plist
@@ -9,6 +9,6 @@
     <key>CFBundleName</key>
     <string>blint</string>
     <key>CFBundleVersion</key>
-    <string>2.2.0</string>
+    <string>2.2.1</string>
 </dict>
 </plist>
diff --git a/blint/analysis.py b/blint/analysis.py
@@ -21,7 +21,7 @@
                           check_virtual_size, check_authenticode,
                           check_dll_characteristics, check_codesign,
                           check_trust_info)
-from blint.config import PII_WORDS, get_int_from_env
+from blint.config import FIRST_STAGE_WORDS, PII_WORDS, get_int_from_env
 from blint.logger import LOG, console
 from blint.utils import (create_findings_table, is_fuzzable_name, print_findings_table)
 
@@ -54,12 +54,20 @@
 review_symbols_dict = defaultdict(list)
 review_imports_dict = defaultdict(list)
 review_entries_dict = defaultdict(list)
-review_rules_cache = {"PII_READ": {
-    "title": "Detect PII Read Operations",
-    "summary": "Can Retrieve Sensitive PII data",
-    "description": "Contains logic to retrieve sensitive data such as names, email, passwords etc.",
-    "patterns": PII_WORDS
-}}
+review_rules_cache = {
+    "PII_READ": {
+        "title": "Detect PII Read Operations",
+        "summary": "Can Retrieve Sensitive PII data",
+        "description": "Contains logic to retrieve sensitive data such as names, email, passwords etc.",
+        "patterns": PII_WORDS
+    },
+    "LOADER_SYMBOLS": {
+        "title": "Detect Initial Loader",
+        "summary": "Behaves like a loader",
+        "description": "The binary behaves like a loader by downloading and executing additional payloads.",
+        "patterns": FIRST_STAGE_WORDS
+    },
+}
 
 # Debug mode
 DEBUG_MODE = os.getenv("SCAN_DEBUG_MODE") == "debug"
@@ -309,7 +317,10 @@ def print_reviews_table(reviews, files):
 def json_serializer(obj):
     """JSON serializer to help serialize problematic types such as bytes"""
     if isinstance(obj, bytes):
-        return obj.decode('utf-8')
+        try:
+            return obj.decode('utf-8')
+        except UnicodeDecodeError:
+            return ""
 
     return obj
 
@@ -494,7 +505,7 @@ def run_review(self, metadata):
                 or self.review_entries_list
         ):
             return self._review_lists(metadata)
-        return {}
+        return self._review_loader_symbols(metadata)
 
     def _review_lists(self, metadata):
         """
@@ -516,6 +527,7 @@ def _review_lists(self, metadata):
         if self.review_entries_list:
             self._review_entries(metadata)
         self._review_pii(metadata)
+        self._review_loader_symbols(metadata)
         return self.results
 
     def _review_imports(self, metadata):
@@ -562,6 +574,22 @@ def _review_pii(self, metadata):
             results["PII_READ"].append({"pattern": e, "function": e})
         self.results |= results
 
+    def _review_loader_symbols(self, metadata):
+        """
+        Reviews loader symbols.
+
+        Args:
+            metadata (dict): The metadata to review.
+
+        Returns:
+            dict: The results of the review.
+        """
+        entries_list = [f.get("name", "") for f in metadata.get("first_stage_symbols", [])]
+        results = defaultdict(list)
+        for e in entries_list[0:EVIDENCE_LIMIT]:
+            results["LOADER_SYMBOLS"].append({"pattern": e, "function": e})
+        self.results |= results
+
     def _review_symbols_exe(self, metadata):
         """
         Reviews symbols in the metadata.

diff --git a/blint/binary.py b/blint/binary.py
@@ -9,7 +9,7 @@
 
 import lief
 
-from blint.config import PII_WORDS, get_float_from_env, get_int_from_env
+from blint.config import FIRST_STAGE_WORDS, PII_WORDS, get_float_from_env, get_int_from_env
 from blint.logger import DEBUG, LOG
 from blint.utils import camel_to_snake, calculate_entropy, check_secret, cleanup_dict_lief_errors, decode_base64
 
@@ -270,7 +270,7 @@ def parse_strings(parsed_obj):
                     if (entropy and (entropy > MIN_ENTROPY or len(s) > MIN_LENGTH)) or secret_type:
                         strings_list.append(
                             {
-                                "value": (decode_base64(s) if s.endswith("==") else s),
+                                "value": decode_base64(s) if s.endswith("==") else s,
                                 "entropy": entropy,
                                 "secret_type": secret_type,
                             }
@@ -788,7 +788,6 @@ def add_elf_metadata(exe_file, metadata, parsed_obj):
         metadata["has_runpath"] = False
     elif runpath:
         metadata["has_runpath"] = True
-    # This is getting renamed to symtab_symbols in lief 0.15.0
     symtab_symbols = parsed_obj.symtab_symbols
     metadata["static"] = bool(symtab_symbols and not isinstance(symtab_symbols, lief.lief_errors))
     dynamic_entries = parsed_obj.dynamic_entries
@@ -797,6 +796,10 @@ def add_elf_metadata(exe_file, metadata, parsed_obj):
     metadata["notes"] = parse_notes(parsed_obj)
     metadata["strings"] = parse_strings(parsed_obj)
     metadata["symtab_symbols"], exe_type = parse_symbols(symtab_symbols)
+    rdata_section = parsed_obj.get_section(".rodata")
+    text_section = parsed_obj.get_section(".text")
+    if not metadata["symtab_symbols"]:
+        add_elf_rdata_symbols(metadata, rdata_section, text_section)
     if exe_type:
         metadata["exe_type"] = exe_type
     metadata["dynamic_symbols"], exe_type = parse_symbols(parsed_obj.dynamic_symbols)
@@ -1114,10 +1117,16 @@ def add_pe_metadata(exe_file: str, metadata: dict, parsed_obj: lief.PE.Binary):
                     break
         rdata_section = parsed_obj.get_section(".rdata")
         text_section = parsed_obj.get_section(".text")
-        if not rdata_section and text_section:
-            rdata_section = text_section
-        if (not metadata["symtab_symbols"] or metadata["exe_type"] != "gobinary") and rdata_section:
-            add_pe_rdata_symbols(metadata, rdata_section)
+        # If there are no .rdata and .text section, then attempt to look for two alphanumeric sections
+        if not rdata_section and not text_section:
+            for section in parsed_obj.sections:
+                if str(section.name).removeprefix(".").isalnum():
+                    if not rdata_section:
+                        rdata_section = section
+                    else:
+                        text_section = section
+        if rdata_section or text_section:
+            add_pe_rdata_symbols(metadata, rdata_section, text_section)
         metadata["exports"] = parse_pe_exports(parsed_obj.get_export())
         metadata["functions"] = parse_functions(parsed_obj.functions)
         metadata["ctor_functions"] = parse_functions(parsed_obj.ctor_functions)
@@ -1247,33 +1256,39 @@ def add_pe_optional_headers(metadata, optional_header):
     return metadata
 
 
-def add_pe_rdata_symbols(metadata, rdata_section: lief.PE.Section):
+def add_pe_rdata_symbols(metadata, rdata_section: lief.PE.Section, text_section: lief.PE.Section):
     """Adds PE rdata symbols to the metadata dictionary.
 
     Args:
         metadata: The dictionary to store the metadata.
         rdata_section: .rdata section of the PE binary.
+        text_section: .text section of the PE binary.
 
     Returns:
         The updated metadata dictionary.
     """
-    if not rdata_section or not rdata_section.content:
-        return metadata
+    file_extns_from_rdata = r".*\.(go|s|dll|exe|pdb)"
     rdata_symbols = set()
     pii_symbols = []
+    first_stage_symbols = []
     for pii in PII_WORDS:
-        for vari in (f"get{pii}", f"get_{camel_to_snake(pii)}"):
-            if rdata_section.search_all(vari):
+        for vari in (f"get{pii}", f"get_{pii}", f"get_{camel_to_snake(pii)}", f"Get{pii}"):
+            if (rdata_section and rdata_section.search_all(vari)) or (text_section and text_section.search_all(vari)):
                 pii_symbols.append(
                     {"name": vari.lower(), "type": "FUNCTION", "is_function": True, "is_imported": False})
                 continue
-    str_content = codecs.decode(rdata_section.content.tobytes("A"), encoding="utf-8", errors="ignore")
+    for sw in FIRST_STAGE_WORDS:
+        if (rdata_section and rdata_section.search_all(sw)) or (text_section and text_section.search_all(sw)):
+            first_stage_symbols.append(
+                {"name": sw, "type": "FUNCTION", "is_function": True, "is_imported": True})
+    str_content = codecs.decode(rdata_section.content.tobytes("A"), encoding="utf-8",
+                                errors="ignore") if rdata_section and rdata_section.content else ""
     for block in str_content.split(" "):
-        if "runtime." in block or "internal/" in block or ".go" in block or ".dll" in block:
+        if "runtime." in block or "internal/" in block or re.match(file_extns_from_rdata, block):
             if ".go" in block:
                 metadata["exe_type"] = "gobinary"
             for asym in block.split("\x00"):
-                if re.match(r".*\.(go|s|dll)$", asym):
+                if re.match(file_extns_from_rdata + "$", asym):
                     rdata_symbols.add(asym)
     if not metadata["symtab_symbols"]:
         metadata["symtab_symbols"] = []
@@ -1285,7 +1300,31 @@ def add_pe_rdata_symbols(metadata, rdata_section: lief.PE.Section):
             "is_imported": True
         } for s in sorted(rdata_symbols)
     ]
-    metadata["pii_symbols"] = pii_symbols
+    if pii_symbols:
+        metadata["pii_symbols"] = pii_symbols
+    if first_stage_symbols:
+        metadata["first_stage_symbols"] = first_stage_symbols
+    return metadata
+
+
+def add_elf_rdata_symbols(metadata, rdata_section: lief.PE.Section, text_section: lief.PE.Section):
+    """Adds ELF rdata symbols to the metadata dictionary.
+
+    Args:
+        metadata: The dictionary to store the metadata.
+        rdata_section: .data section of the ELF binary.
+        text_section: .text section of the ELF binary.
+
+    Returns:
+        The updated metadata dictionary.
+    """
+    first_stage_symbols = []
+    for sw in FIRST_STAGE_WORDS:
+        if (rdata_section and rdata_section.search_all(sw)) or (text_section and text_section.search_all(sw)):
+            first_stage_symbols.append(
+                {"name": sw, "type": "FUNCTION", "is_function": True, "is_imported": True})
+    if first_stage_symbols:
+        metadata["first_stage_symbols"] = first_stage_symbols
     return metadata
 
 

diff --git a/blint/config.py b/blint/config.py
@@ -1230,7 +1230,7 @@
     "email": [re.compile(r"(?<=mailto:)[a-zA-Z0-9_.+-]+@[a-zA-Z0-9-]+.[a-zA-Z0-9.-]+")],
     "ip": [
         re.compile(
-            r"^((25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9]).){3}(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9])$"
+            r"^((25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9]).){3}(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9])(:[0-9]+)?$"
         )
     ],
 }
@@ -1268,6 +1268,7 @@ def get_int_from_env(name, default):
     return int(get_float_from_env(name, default))
 
 
+# PII related symbols
 PII_WORDS = (
     "FirstName",
     "LastName",
@@ -1293,3 +1294,43 @@ def get_int_from_env(name, default):
     "AgentStatus",
     "LastLoginTime"
 )
+
+# Some symbols to look for in a first-stage payload
+FIRST_STAGE_WORDS = (
+    "System.ServiceProcess",
+    "System.IO.Compression",
+    "System.Reflection.Emit",
+    "ICryptoTransform",
+    "LoadAssembly",
+    "GetEncodedData",
+    "add_AssemblyResolve",
+    "CreateDecryptor",
+    "GetExecutingAssembly",
+    "GetModules",
+    "get_IsFamilyOrAssembly",
+    "/proc/%d/cmdline",
+    "/proc/%s/exe",
+    "/proc/self/exe",
+    "/proc/net/route",
+    "/etc/resolv.conf",
+    "/usr/lib/systemd/systemd",
+    "/usr/compress/bin/",
+    "/usr/libexec/openssh/sftp-server",
+    "/usr/sbin/reboot",
+    "/usr/bin/reboot",
+    "/usr/sbin/shutdown",
+    "/usr/bin/shutdown",
+    "/usr/sbin/poweroff",
+    "/usr/bin/poweroff",
+    "/usr/sbin/halt",
+    "/usr/bin/halt",
+    "virtualboxemulunit",
+    "virtualboximportunit",
+    "virtualboxunit",
+    "qvirtualboxglobalsunit",
+    "Uvirtualboxdisasm",
+    "loaderx86.dll",
+    "ZwProtectVirtualMemory",
+    "shlwapi.dll",
+    "DeleteCriticalSection"
+)
diff --git a/blint/data/annotations/review_exe_go.yml b/blint/data/annotations/review_exe_go.yml
@@ -4,6 +4,7 @@ group: EXE_REVIEWS
 exe_type:
   - gobinary
   - x86_64-executable
+  - x86_64-exec
 rules:
   - id: FILE_IOUTIL
     title: IO util functions used

diff --git a/blint/data/annotations/review_imports_pe.yml b/blint/data/annotations/review_imports_pe.yml
@@ -41,6 +41,11 @@ rules:
       - CreateSocketHandle
       - CreateThread
       - CreateThreadpool
+      - GetCurrentThread
+      - GetThreadLocale
+      - ExitProcess
+      - FreeLibrary
+      - ExitThread
   - id: PROCESS_STATUS_API
     title: Process status api used
     summary: Queries about processes and device drivers
@@ -464,6 +469,8 @@ rules:
       - CreateHardLink
       - BackupEventLog
       - CreateFileMapping
+      - SetFileAttributes
+      - SetFilePointer
   - id: WIN_BCRYPT_API
     title: Win bcrypt api functions used
     summary: Can Encrypt Files
@@ -1973,6 +1980,14 @@ rules:
     description: |
       Shell Windows API allows applications to access functions provided by the operating system shell, and to change and enhance it.
     patterns:
+      - DllInstall
+      - GetProcessReference
+      - ParseURLA
+      - ParseURLW
+      - PathFindFileName
+      - PathFindOnPath
+      - PathIsSystemFolder
+      - PathIsUNCServerShare
       - FindExecutableA
       - FindExecutableW
       - InitNetworkAddressControl
@@ -1994,6 +2009,30 @@ rules:
       - GetManagedApplications
       - InstallApplication
       - UninstallApplication
+      - SHCreateThread
+      - SHCreateStreamOnFile
+      - SHOpenRegStream
+      - SHRegCreateUSKey
+      - SHQueryValueEx
+      - SHQueryInfoKey
+      - UrlCreateFromPath
+      - PathMatchSpec
+      - SHBindToFolderIDListParent
+      - SHBrowseForFolder
+      - SHCreateDefaultContextMenu
+      - SHCreateShellItem
+      - SHFormatDrive
+      - SHGetDesktopFolder
+      - SHGetFolderPath
+      - SHGetFolderPathAndSubDir
+      - SHGetKnownFolder
+      - SHGetSpecialFolderLocation
+      - SHILCreateFromPath
+      - SHPathPrepareForWrite
+      - SHSetKnownFolderPath
+      - SignalFileOpen
+      - Win32DeleteFile
+      - shlwapi.dll
   - id: KERNEL_API
     title: Kernel api functions used
     summary: Manipulates Windows Kernel & Drivers

diff --git a/blint/data/annotations/review_monero_go.yml b/blint/data/annotations/review_monero_go.yml
@@ -4,6 +4,7 @@ group: SYMBOL_REVIEWS
 exe_type: 
   - gobinary
   - x86_64-executable
+  - x86_64-exec
 rules:
   - id: MONERO_API_GO
     title: Detect use of Monero wallet

diff --git a/blint/data/annotations/review_rootkits_win.yml b/blint/data/annotations/review_rootkits_win.yml
@@ -3,6 +3,7 @@ text: Review for Windows rootkits
 group: METHOD_REVIEWS
 exe_type:
   - x86_64-executable
+  - x86_64-exec
   - PE32
   - PE64
 rules:
@@ -67,7 +68,6 @@ rules:
       - BeIsStringNull
       - BeIsStringTerminated
       - BeUnSupportedFunction
-
   - id: NIDHOGG
     title: Detect Nidhogg
     summary: Provides Tools for Gaining Privileged Access and Injecting Malicious Code