... | ... |
@@ -15,7 +15,6 @@ Table Of Contents |
15 | 15 |
5. [ClamAV Developer Tips and Tricks](UserManual/development.md) |
16 | 16 |
6. [Build \[lib\]ClamAV Into Your Programs](UserManual/libclamav.md) |
17 | 17 |
7. [Writing ClamAV Signatures](UserManual/Signatures.md) |
18 |
-8. [Writing ClamAV Phishing Signatures](UserManual/PhishSigs.md) |
|
19 | 18 |
|
20 | 19 |
----- |
21 | 20 |
|
22 | 21 |
deleted file mode 100644 |
... | ... |
@@ -1,89 +0,0 @@ |
1 |
-# ClamAV File Types |
|
2 |
- |
|
3 |
-ClamAV maintains it's own file typing format and assigns these types using either: |
|
4 |
- |
|
5 |
-- Evaluation of a unique sequence of bytes at the start of a file (file type magic). |
|
6 |
-- File type indicators when parsing container files. |
|
7 |
- - For example: |
|
8 |
- CL_TYPE_SCRIPT may be assigned to data contained in a PDF when the PDF indicates that a stream of bytes is "Javascript" |
|
9 |
-- File type determination based on the names or characteristics contained within the file. |
|
10 |
- - For example: |
|
11 |
- CL_TYPE_OOXML_WORD may be assigned to a Zip file containing files with specific names. |
|
12 |
- |
|
13 |
-## CL_TYPE's |
|
14 |
- |
|
15 |
-ClamAV Types are prefixed with `CL_TYPE_`. The following is an exhaustive list of all current CL_TYPE's. |
|
16 |
- |
|
17 |
-| CL_TYPE | Description | |
|
18 |
-|------------------------|-----------------------------------------------------------------------------------| |
|
19 |
-| `CL_TYPE_7Z` | 7-Zip Archive | |
|
20 |
-| `CL_TYPE_7ZSFX` | Self-Extracting 7-Zip Archive | |
|
21 |
-| `CL_TYPE_APM` | Disk Image - Apple Partition Map | |
|
22 |
-| `CL_TYPE_ARJ` | ARJ Archive | |
|
23 |
-| `CL_TYPE_ARJSFX` | Self-Extracting ARJ Archive | |
|
24 |
-| `CL_TYPE_AUTOIT` | AutoIt Automation Executable | |
|
25 |
-| `CL_TYPE_BINARY_DATA` | binary data | |
|
26 |
-| `CL_TYPE_BINHEX` | BinHex Macintosh 7-bit ASCII email attachment encoding | |
|
27 |
-| `CL_TYPE_BZ` | BZip Compressed File | |
|
28 |
-| `CL_TYPE_CABSFX` | Self-Extracting Microsoft CAB Archive | |
|
29 |
-| `CL_TYPE_CPIO_CRC` | CPIO Archive (CRC) | |
|
30 |
-| `CL_TYPE_CPIO_NEWC` | CPIO Archive (NEWC) | |
|
31 |
-| `CL_TYPE_CPIO_ODC` | CPIO Archive (ODC) | |
|
32 |
-| `CL_TYPE_CPIO_OLD` | CPIO Archive (OLD, Little Endian or Big Endian) | |
|
33 |
-| `CL_TYPE_CRYPTFF` | Files encrypted by CryptFF malware | |
|
34 |
-| `CL_TYPE_DMG` | Apple DMG Archive | |
|
35 |
-| `CL_TYPE_ELF` | ELF Executable (Linux/Unix program or library) | |
|
36 |
-| `CL_TYPE_GPT` | Disk Image - GUID Partition Table | |
|
37 |
-| `CL_TYPE_GRAPHICS` | TIFF (Little Endian or Big Endian) | |
|
38 |
-| `CL_TYPE_GZ` | GZip Compressed File | |
|
39 |
-| `CL_TYPE_HTML_UTF16` | Wide-Character / UTF16 encoded HTML | |
|
40 |
-| `CL_TYPE_HTML` | HTML data | |
|
41 |
-| `CL_TYPE_HWP3` | Hangul Word Processor (3.X) | |
|
42 |
-| `CL_TYPE_HWPOLE2` | Hangul Word Processor embedded OLE2 | |
|
43 |
-| `CL_TYPE_INTERNAL` | Internal properties | |
|
44 |
-| `CL_TYPE_ISHIELD_MSI` | Windows Install Shield MSI installer | |
|
45 |
-| `CL_TYPE_ISO9660` | ISO 9660 file system for optical disc media | |
|
46 |
-| `CL_TYPE_JAVA` | Java Class File | |
|
47 |
-| `CL_TYPE_LNK` | Microsoft Windows Shortcut File | |
|
48 |
-| `CL_TYPE_MACHO_UNIBIN` | Universal Binary/Java Bytecode | |
|
49 |
-| `CL_TYPE_MACHO` | Apple/NeXTSTEP Mach-O Executable file format | |
|
50 |
-| `CL_TYPE_MAIL` | Email file | |
|
51 |
-| `CL_TYPE_MBR` | Disk Image - Master Boot Record | |
|
52 |
-| `CL_TYPE_MHTML` | MHTML Saved Web Page | |
|
53 |
-| `CL_TYPE_MSCAB` | Microsoft CAB Archive | |
|
54 |
-| `CL_TYPE_MSCHM` | Microsoft CHM help archive | |
|
55 |
-| `CL_TYPE_MSEXE` | Microsoft EXE / DLL Executable file | |
|
56 |
-| `CL_TYPE_MSOLE2` | Microsoft OLE2 Container file | |
|
57 |
-| `CL_TYPE_MSSZDD` | Microsoft Compressed EXE | |
|
58 |
-| `CL_TYPE_NULSFT` | NullSoft Scripted Installer program | |
|
59 |
-| `CL_TYPE_OLD_TAR` | TAR archive (old) | |
|
60 |
-| `CL_TYPE_OOXML_HWP` | Hangul Office Open Word Processor (5.X) | |
|
61 |
-| `CL_TYPE_OOXML_PPT` | Microsoft Office Open XML PowerPoint | |
|
62 |
-| `CL_TYPE_OOXML_WORD` | Microsoft Office Open Word 2007+ | |
|
63 |
-| `CL_TYPE_OOXML_XL` | Microsoft Office Open Excel 2007+ | |
|
64 |
-| `CL_TYPE_PART_HFSPLUS` | Apple HFS+ partition | |
|
65 |
-| `CL_TYPE_PDF` | Adobe PDF document | |
|
66 |
-| `CL_TYPE_POSIX_TAR` | TAR archive | |
|
67 |
-| `CL_TYPE_PS` | Postscript | |
|
68 |
-| `CL_TYPE_RAR` | RAR Archive | |
|
69 |
-| `CL_TYPE_RARSFX` | Self-Extracting RAR Archive | |
|
70 |
-| `CL_TYPE_RIFF` | Resource Interchange File Format container formatted file | |
|
71 |
-| `CL_TYPE_RTF` | Rich Text Format document | |
|
72 |
-| `CL_TYPE_SCRENC` | Files encrypted by ScrEnc malware | |
|
73 |
-| `CL_TYPE_SCRIPT` | Generic type for scripts that don't have their own type (Javascript, Python, etc) | |
|
74 |
-| `CL_TYPE_SIS` | Symbian OS Software Installation Script Archive | |
|
75 |
-| `CL_TYPE_SWF` | Adobe Flash File (LZMA, Zlib, or uncompressed) | |
|
76 |
-| `CL_TYPE_TEXT_ASCII` | ASCII text | |
|
77 |
-| `CL_TYPE_TEXT_UTF16BE` | UTF-16BE text | |
|
78 |
-| `CL_TYPE_TEXT_UTF16LE` | UTF-16LE text | |
|
79 |
-| `CL_TYPE_TEXT_UTF8` | UTF-8 text | |
|
80 |
-| `CL_TYPE_TNEF` | Microsoft Outlook & Exchange email attachment format | |
|
81 |
-| `CL_TYPE_UUENCODED` | UUEncoded (Unix-to-Unix) binary file (Unix email attachment format) | |
|
82 |
-| `CL_TYPE_XAR` | XAR Archive | |
|
83 |
-| `CL_TYPE_XDP` | Adobe XDP - Embedded PDF | |
|
84 |
-| `CL_TYPE_XML_HWP` | Hangul Word Processor XML (HWPML) Document | |
|
85 |
-| `CL_TYPE_XML_WORD` | Microsoft Word 2003 XML Document | |
|
86 |
-| `CL_TYPE_XML_XL` | Microsoft Excel 2003 XML Document | |
|
87 |
-| `CL_TYPE_XZ` | XZ Archive | |
|
88 |
-| `CL_TYPE_ZIP` | Zip Archive | |
|
89 |
-| `CL_TYPE_ZIPSFX` | Self-Extracting Zip Archive | |
9 | 9 |
deleted file mode 100644 |
... | ... |
@@ -1,682 +0,0 @@ |
1 |
-# PhishSigs |
|
2 |
- |
|
3 |
-Table of Contents |
|
4 |
-- [PhishSigs](#phishsigs) |
|
5 |
-- [Database file format](#database-file-format) |
|
6 |
- - [PDB format](#pdb-format) |
|
7 |
- - [GDB format](#gdb-format) |
|
8 |
- - [WDB format](#wdb-format) |
|
9 |
- - [Hints](#hints) |
|
10 |
- - [Examples of PDB signatures](#examples-of-pdb-signatures) |
|
11 |
- - [Examples of WDB signatures](#examples-of-wdb-signatures) |
|
12 |
- - [Example for how the URL extractor works](#example-for-how-the-url-extractor-works) |
|
13 |
- - [How matching works](#how-matching-works) |
|
14 |
- - [RealURL, displayedURL concatenation](#realurl-displayedurl-concatenation) |
|
15 |
- - [What happens when a match is found](#what-happens-when-a-match-is-found) |
|
16 |
- - [Extraction of realURL, displayedURL from HTML tags](#extraction-of-realurl-displayedurl-from-html-tags) |
|
17 |
- - [Example](#example) |
|
18 |
- - [Simple patterns](#simple-patterns) |
|
19 |
- - [Regular expressions](#regular-expressions) |
|
20 |
- - [Flags](#flags) |
|
21 |
-- [Introduction to regular expressions](#introduction-to-regular-expressions) |
|
22 |
- - [Special characters](#special-characters) |
|
23 |
- - [Character classes](#character-classes) |
|
24 |
- - [Escaping](#escaping) |
|
25 |
- - [Alternation](#alternation) |
|
26 |
- - [Optional matching, and repetition](#optional-matching-and-repetition) |
|
27 |
- - [Groups](#groups) |
|
28 |
-- [How to create database files](#how-to-create-database-files) |
|
29 |
- - [How to create and maintain the whitelist (daily.wdb)](#how-to-create-and-maintain-the-whitelist-dailywdb) |
|
30 |
- - [How to create and maintain the domainlist (daily.pdb)](#how-to-create-and-maintain-the-domainlist-dailypdb) |
|
31 |
- - [Dealing with false positives, and undetected phishing mails](#dealing-with-false-positives-and-undetected-phishing-mails) |
|
32 |
- - [False positives](#false-positives) |
|
33 |
- - [Undetected phish mails](#undetected-phish-mails) |
|
34 |
- |
|
35 |
-# Database file format |
|
36 |
- |
|
37 |
-## PDB format |
|
38 |
- |
|
39 |
-This file contains urls/hosts that are target of phishing attempts. It |
|
40 |
-contains lines in the following format: |
|
41 |
- |
|
42 |
-``` |
|
43 |
- R[Filter]:RealURL:DisplayedURL[:FuncLevelSpec] |
|
44 |
- H[Filter]:DisplayedHostname[:FuncLevelSpec] |
|
45 |
-``` |
|
46 |
- |
|
47 |
-- `R` |
|
48 |
- |
|
49 |
- regular expression, for the concatenated URL |
|
50 |
- |
|
51 |
-- `H` |
|
52 |
- |
|
53 |
- matches the `DisplayedHostname` as a simple pattern (literally, no regular expression) |
|
54 |
- |
|
55 |
- - the pattern can match either the full hostname |
|
56 |
- |
|
57 |
- - or a subdomain of the specified hostname |
|
58 |
- |
|
59 |
- - to avoid false matches in case of subdomain matches, the engine checks that there is a dot(`.`) or a space(` `) before the matched portion |
|
60 |
- |
|
61 |
-- `Filter` |
|
62 |
- |
|
63 |
- is ignored for R and H for compatibility reasons |
|
64 |
- |
|
65 |
-- `RealURL` |
|
66 |
- |
|
67 |
- is the URL the user is sent to, example: *href* attribute of an html anchor (*\<a\> tag*) |
|
68 |
- |
|
69 |
-- `DisplayedURL` |
|
70 |
- |
|
71 |
- is the URL description displayed to the user, where its *claimed* they are sent, example: contents of an html anchor (*\<a\> tag*) |
|
72 |
- |
|
73 |
-- `DisplayedHostname` |
|
74 |
- |
|
75 |
- is the hostname portion of the DisplayedURL |
|
76 |
- |
|
77 |
-- `FuncLevelSpec` |
|
78 |
- |
|
79 |
- an (optional) functionality level, 2 formats are possible: |
|
80 |
- |
|
81 |
- - `minlevel` all engines having functionality level \>= `minlevel` will load this line |
|
82 |
- |
|
83 |
- - `minlevel-maxlevel` engines with functionality level \(>=\) `minlevel`, and \(<\) `maxlevel` will load this line |
|
84 |
- |
|
85 |
-## GDB format |
|
86 |
- |
|
87 |
-This file contains URL hashes in the following format: |
|
88 |
- |
|
89 |
- S:P:HostPrefix[:FuncLevelSpec] |
|
90 |
- S:F:Sha256hash[:FuncLevelSpec] |
|
91 |
- S1:P:HostPrefix[:FuncLevelSpec] |
|
92 |
- S1:F:Sha256hash[:FuncLevelSpec] |
|
93 |
- S2:P:HostPrefix[:FuncLevelSpec] |
|
94 |
- S2:F:Sha256hash[:FuncLevelSpec] |
|
95 |
- S:W:Sha256hash[:FuncLevelSpec] |
|
96 |
- |
|
97 |
-- `S:` |
|
98 |
- |
|
99 |
- These are hashes for Google Safe Browsing - malware sites, and should not be used for other purposes. |
|
100 |
- |
|
101 |
-- `S2:` |
|
102 |
- |
|
103 |
- These are hashes for Google Safe Browsing - phishing sites, and should not be used for other purposes. |
|
104 |
- |
|
105 |
-- `S1:` |
|
106 |
- |
|
107 |
- Hashes for blacklisting phishing sites. Virus name: Phishing.URL.Blacklisted |
|
108 |
- |
|
109 |
-- `S:W:` |
|
110 |
- |
|
111 |
- Locally whitelisted hashes. |
|
112 |
- |
|
113 |
-- `HostPrefix` |
|
114 |
- |
|
115 |
- 4-byte prefix of the sha256 hash of the last 2 or 3 components of the hostname. If prefix doesn’t match, no further lookups are performed. |
|
116 |
- |
|
117 |
-- `Sha256hash` |
|
118 |
- |
|
119 |
- sha256 hash of the canonicalized URL, or a sha256 hash of its prefix/suffix according to the Google Safe Browsing “Performing Lookups” rules. There should be a corresponding `:P:HostkeyPrefix` entry for the hash to be taken into consideration. |
|
120 |
- |
|
121 |
-To see which hash/URL matched, look at the `clamscan --debug` output, and look for the following strings: `Looking up hash`, `prefix matched`, and `Hash matched`. Local whitelisting of .gdb entries can be done by creating a local.gdb file, and adding a line `S:W:<HASH>`. |
|
122 |
- |
|
123 |
-## WDB format |
|
124 |
- |
|
125 |
-This file contains whitelisted url pairs It contains lines in the following format: |
|
126 |
- |
|
127 |
-``` |
|
128 |
- X:RealURL:DisplayedURL[:FuncLevelSpec] |
|
129 |
- M:RealHostname:DisplayedHostname[:FuncLevelSpec] |
|
130 |
-``` |
|
131 |
- |
|
132 |
-- `X` |
|
133 |
- |
|
134 |
- regular expression, for the *entire URL*, not just the hostname |
|
135 |
- |
|
136 |
- - The regular expression is by default anchored to start-of-line and end-of-line, as if you have used `^RegularExpression$` |
|
137 |
- |
|
138 |
- - A trailing `/` is automatically added both to the regex, and the input string to avoid false matches |
|
139 |
- |
|
140 |
- - The regular expression matches the *concatenation* of the RealURL, a colon(`:`), and the DisplayedURL as a single string. It doesn’t separately match RealURL and DisplayedURL\! |
|
141 |
- |
|
142 |
-- `M` |
|
143 |
- |
|
144 |
- matches hostname, or subdomain of it, see notes for H above |
|
145 |
- |
|
146 |
-## Hints |
|
147 |
- |
|
148 |
-- empty lines are ignored |
|
149 |
- |
|
150 |
-- the colons are mandatory |
|
151 |
- |
|
152 |
-- Don’t leave extra spaces on the end of a line\! |
|
153 |
- |
|
154 |
-- if any of the lines don’t conform to this format, clamav will abort with a Malformed Database Error |
|
155 |
- |
|
156 |
-- see section [Extraction-of-realURL](#Extraction-of-realURL,-displayedURL-from-HTML-tags) for more details on realURL/displayedURL |
|
157 |
- |
|
158 |
-## Examples of PDB signatures |
|
159 |
- |
|
160 |
-To check for phishing mails that target amazon.com, or subdomains of |
|
161 |
-amazon.com: |
|
162 |
- |
|
163 |
-``` |
|
164 |
- H:amazon.com |
|
165 |
-``` |
|
166 |
- |
|
167 |
-To do the same, but for amazon.co.uk: |
|
168 |
- |
|
169 |
-``` |
|
170 |
- H:amazon.co.uk |
|
171 |
-``` |
|
172 |
- |
|
173 |
-To limit the signatures to certain engine versions: |
|
174 |
- |
|
175 |
-``` |
|
176 |
- H:amazon.co.uk:20-30 |
|
177 |
- H:amazon.co.uk:20- |
|
178 |
- H:amazon.co.uk:0-20 |
|
179 |
-``` |
|
180 |
- |
|
181 |
-First line: engine versions 20, 21, ..., 29 can load it |
|
182 |
- |
|
183 |
-Second line: engine versions \>= 20 can load it |
|
184 |
- |
|
185 |
-Third line: engine versions \< 20 can load it |
|
186 |
- |
|
187 |
-In a real situation, you’d probably use the second form. A situation like that would be if you are using a feature of the signatures not available in earlier versions, or if earlier versions have bugs with your signature. Its neither case here, the above examples are for illustrative purposes only. |
|
188 |
- |
|
189 |
-## Examples of WDB signatures |
|
190 |
- |
|
191 |
-To allow amazon’s country specific domains and amazon.com, to mix domain names in DisplayedURL, and RealURL: |
|
192 |
- |
|
193 |
- X:.+\.amazon\.(at|ca|co\.uk|co\.jp|de|fr)([/?].*)?:.+\.amazon\.com([/?].*)?:17- |
|
194 |
- |
|
195 |
-Explanation of this signature: |
|
196 |
- |
|
197 |
-- `X:` |
|
198 |
- |
|
199 |
- this is a regular expression |
|
200 |
- |
|
201 |
-- `:17-` |
|
202 |
- |
|
203 |
- load signature only for engines with functionality level \>= 17 (recommended for type X) |
|
204 |
- |
|
205 |
-The regular expression is the following (X:, :17- stripped, and a / appended) |
|
206 |
- |
|
207 |
-``` |
|
208 |
- .+\.amazon\.(at|ca|co\.uk|co\.jp|de|fr)([/?].*)?:.+\.amazon\.com([/?].*)?/ |
|
209 |
-``` |
|
210 |
- |
|
211 |
-Explanation of this regular expression (note that it is a single regular expression, and not 2 regular expressions splitted at the :). |
|
212 |
- |
|
213 |
-- `.+` |
|
214 |
- |
|
215 |
- any subdomain of |
|
216 |
- |
|
217 |
-- `\.amazon\.` |
|
218 |
- |
|
219 |
- domain we are whitelisting (RealURL part) |
|
220 |
- |
|
221 |
-- `(at|ca|co\.uk|co\.jp|de|fr)` |
|
222 |
- |
|
223 |
- country-domains: at, ca, co.uk, co.jp, de, fr |
|
224 |
- |
|
225 |
-- `([/?].*)?` |
|
226 |
- |
|
227 |
- recomended way to end real url part of whitelist, this protects against embedded URLs (evilurl.example.com/amazon.co.uk/) |
|
228 |
- |
|
229 |
-- `:` |
|
230 |
- |
|
231 |
- RealURL and DisplayedURL are concatenated via a :, so match a literal : here |
|
232 |
- |
|
233 |
-- `.+` |
|
234 |
- |
|
235 |
- any subdomain of |
|
236 |
- |
|
237 |
-- `\.amazon\.com` |
|
238 |
- |
|
239 |
- whitelisted DisplayedURL |
|
240 |
- |
|
241 |
-- `([/?].*)?` |
|
242 |
- |
|
243 |
- recommended way to end displayed url part, to protect against embedded URLs |
|
244 |
- |
|
245 |
-- `/` |
|
246 |
- |
|
247 |
- automatically added to further protect against embedded URLs |
|
248 |
- |
|
249 |
-When you whitelist an entry make sure you check that both domains are owned by the same entity. What this whitelist entry allows is: Links claiming to point to amazon.com (DisplayedURL), but really go to country-specific domain of amazon (RealURL). |
|
250 |
- |
|
251 |
-## Example for how the URL extractor works |
|
252 |
- |
|
253 |
-Consider the following HTML file: |
|
254 |
- |
|
255 |
-```html |
|
256 |
- <html> |
|
257 |
- <a href="http://1.realurl.example.com/"> |
|
258 |
- 1.displayedurl.example.com |
|
259 |
- </a> |
|
260 |
- <a href="http://2.realurl.example.com"> |
|
261 |
- 2 d<b>i<p>splayedurl.e</b>xa<i>mple.com |
|
262 |
- </a> |
|
263 |
- <a href="http://3.realurl.example.com"> |
|
264 |
- 3.nested.example.com |
|
265 |
- <a href="http://4.realurl.example.com"> |
|
266 |
- 4.displayedurl.example.com |
|
267 |
- </a> |
|
268 |
- </a> |
|
269 |
- <form action="http://5.realurl.example.com"> |
|
270 |
- sometext |
|
271 |
- <img src="http://5.displayedurl.example.com/img0.gif"/> |
|
272 |
- <a href="http://5.form.nested.displayedurl.example.com"> |
|
273 |
- 5.form.nested.link-displayedurl.example.com |
|
274 |
- </a> |
|
275 |
- </form> |
|
276 |
- <a href="http://6.realurl.example.com"> |
|
277 |
- 6.displ |
|
278 |
- <img src="6.displayedurl.example.com/img1.gif"/> |
|
279 |
- ayedurl.example.com |
|
280 |
- </a> |
|
281 |
- <a href="http://7.realurl.example.com"> |
|
282 |
- <iframe src="http://7.displayedurl.example.com"> |
|
283 |
- </a> |
|
284 |
-``` |
|
285 |
- |
|
286 |
-The phishing engine extract the following |
|
287 |
-RealURL/DisplayedURL pairs from it: |
|
288 |
- |
|
289 |
-``` |
|
290 |
- http://1.realurl.example.com/ |
|
291 |
- 1.displayedurl.example.com |
|
292 |
- |
|
293 |
- http://2.realurl.example.com |
|
294 |
- 2displayedurl.example.com |
|
295 |
- |
|
296 |
- http://3.realurl.example.com |
|
297 |
- 3.nested.example.com |
|
298 |
- |
|
299 |
- http://4.realurl.example.com |
|
300 |
- 4.displayedurl.example.com |
|
301 |
- |
|
302 |
- http://5.realurl.example.com |
|
303 |
- http://5.displayedurl.example.com/img0.gif |
|
304 |
- |
|
305 |
- http://5.realurl.example.com |
|
306 |
- http://5.form.nested.displayedurl.example.com |
|
307 |
- |
|
308 |
- http://5.form.nested.displayedurl.example.com |
|
309 |
- 5.form.nested.link-displayedurl.example.com |
|
310 |
- |
|
311 |
- http://6.realurl.example.com |
|
312 |
- 6.displayedurl.example.com |
|
313 |
- |
|
314 |
- http://6.realurl.example.com |
|
315 |
- 6.displayedurl.example.com/img1.gif |
|
316 |
-``` |
|
317 |
- |
|
318 |
-## How matching works |
|
319 |
- |
|
320 |
-### RealURL, displayedURL concatenation |
|
321 |
- |
|
322 |
-The phishing detection module processes pairs of RealURL/DisplayedURL. Matching against daily.wdb is done as follows: the realURL is concatenated with a `:`, and with the DisplayedURL, then that *line* is matched against the lines in daily.wdb/daily.pdb |
|
323 |
- |
|
324 |
-So if you have this line in daily.wdb: |
|
325 |
- |
|
326 |
- M:www.google.ro:www.google.com |
|
327 |
- |
|
328 |
-and this href: `<a href='http://www.google.ro'>www.google.com</a>` then it will be whitelisted, but: `<a href='http://images.google.com'>www.google.com</a>` will not. |
|
329 |
- |
|
330 |
-### What happens when a match is found |
|
331 |
- |
|
332 |
-In the case of the whitelist, a match means that the RealURL/DisplayedURL combination is considered clean, and no further checks are performed on it. |
|
333 |
- |
|
334 |
-In the case of the domainlist, a match means that the RealURL/displayedURL is going to be checked for phishing attempts. |
|
335 |
- |
|
336 |
-Furthermore you can restrict what checks are to be performed by specifying the 3-digit hexnumber. |
|
337 |
- |
|
338 |
-### Extraction of realURL, displayedURL from HTML tags |
|
339 |
- |
|
340 |
-The html parser extracts pairs of realURL/displayedURL based on the following rules. |
|
341 |
- |
|
342 |
-In version 0.93: After URLs have been extracted, they are normalized, and cut after the hostname. `http://test.example.com/path/somecgi?queryparameters` becomes `http://test.example.com/` |
|
343 |
- |
|
344 |
-- `a` |
|
345 |
- |
|
346 |
- (anchor) the *href* is the realURL, its *contents* is the displayedURL |
|
347 |
- |
|
348 |
- - contents |
|
349 |
- is the tag-stripped contents of the \<a\> tags, so for example \<b\> tags are stripped (but not their contents) |
|
350 |
- |
|
351 |
- nesting another \<a\> tag withing an \<a\> tag (besides being invalid html) is treated as a \</a\>\<a.. |
|
352 |
- |
|
353 |
-- `form` |
|
354 |
- |
|
355 |
- the *action* attribute is the realURL, and a nested \<a\> tag is the displayedURL |
|
356 |
- |
|
357 |
-- `img/area` |
|
358 |
- |
|
359 |
- if nested within an *\<a\>* tag, the realURL is the *href* of the a tag, and the *src/dynsrc/area* is the displayedURL of the img |
|
360 |
- |
|
361 |
- if nested withing a *form* tag, then the action attribute of the *form* tag is the realURL |
|
362 |
- |
|
363 |
-- `iframe` |
|
364 |
- |
|
365 |
- if nested withing an *\<a\>* tag the *src* attribute is the displayedURL, and the *href* of its parent *a* tag is the realURL |
|
366 |
- |
|
367 |
- if nested withing a *form* tag, then the action attribute of the *form* tag is the realURL |
|
368 |
- |
|
369 |
-### Example |
|
370 |
- |
|
371 |
-Consider this html file: |
|
372 |
- |
|
373 |
-```html |
|
374 |
-<a href=”evilurl”\>www.paypal.com\</a\>* |
|
375 |
- |
|
376 |
-<a href=”evilurl2” title=”www.ebay.com”\>click here to sign |
|
377 |
-in\</a\>* |
|
378 |
- |
|
379 |
-<form action=”evilurl_form”\>* |
|
380 |
- |
|
381 |
-*Please sign in to \<a href=”cgi.ebay.com”\>Ebay\</a\using this |
|
382 |
-form* |
|
383 |
- |
|
384 |
-<input type=’text’ name=’username’\>Username\</input\>* |
|
385 |
- |
|
386 |
-*....* |
|
387 |
- |
|
388 |
-</form\>* |
|
389 |
- |
|
390 |
-<a href=”evilurl”\>\<img src=”images.paypal.com/secure.jpg”\>\</a\>* |
|
391 |
-``` |
|
392 |
- |
|
393 |
-The resulting realURL/displayedURL pairs will be (note that one tag can generate multiple pairs): |
|
394 |
- |
|
395 |
-- evilurl / www.paypal.com |
|
396 |
- |
|
397 |
-- evilurl2 / click here to sign in |
|
398 |
- |
|
399 |
-- evilurl2 / www.ebay.com |
|
400 |
- |
|
401 |
-- evilurl_form / cgi.ebay.com |
|
402 |
- |
|
403 |
-- cgi.ebay.com / Ebay |
|
404 |
- |
|
405 |
-- evilurl / image.paypal.com/secure.jpg |
|
406 |
- |
|
407 |
-## Simple patterns |
|
408 |
- |
|
409 |
-Simple patterns are matched literally, i.e. if you say: |
|
410 |
- |
|
411 |
-``` |
|
412 |
-www.google.com |
|
413 |
-``` |
|
414 |
- |
|
415 |
-it is going to match *www.google.com*, and only that. The *. (dot)* character has no special meaning (see the section on regexes [\[sec:Regular-expressions\]](#sec:Regular-expressions) for how the *.(dot)* character behaves there) |
|
416 |
- |
|
417 |
-## Regular expressions |
|
418 |
- |
|
419 |
-POSIX regular expressions are supported, and you can consider that internally it is wrapped by *^*, and *$.* In other words, this means that the regular expression has to match the entire concatenated (see section [RealURL,-displayedURL-concatenation](#RealURL,-displayedURL-concatenation) for details on concatenation) url. |
|
420 |
- |
|
421 |
-It is recomended that you read section [Introduction-to-regular](#Introduction-to-regular) to learn how to write regular expressions, and then come back and read this for hints. |
|
422 |
- |
|
423 |
-Be advised that clamav contains an internal, very basic regex matcher to reduce the load on the regex matching core. Thus it is recomended that you avoid using regex syntax not supported by it at the very beginning of regexes (at least the first few characters). |
|
424 |
- |
|
425 |
-Currently the clamav regex matcher supports: |
|
426 |
- |
|
427 |
-- `.` (dot) character |
|
428 |
- |
|
429 |
-- `\(\backslash\)` (escaping special characters) |
|
430 |
- |
|
431 |
-- `|` (pipe) alternatives |
|
432 |
- |
|
433 |
-- `\[\]` (character classes) |
|
434 |
- |
|
435 |
-- `()` (parenthesis for grouping, but no group extraction is performed) |
|
436 |
- |
|
437 |
-- other non-special characters |
|
438 |
- |
|
439 |
-Thus the following are not supported: |
|
440 |
- |
|
441 |
-- `\+` repetition |
|
442 |
- |
|
443 |
-- `\*` repetition |
|
444 |
- |
|
445 |
-- `{}` repetition |
|
446 |
- |
|
447 |
-- backreferences |
|
448 |
- |
|
449 |
-- lookaround |
|
450 |
- |
|
451 |
-- other “advanced” features not listed in the supported list ;) |
|
452 |
- |
|
453 |
-This however shouldn’t discourage you from using the “not directly supported features “, because if the internal engine encounters unsupported syntax, it passes it on to the POSIX regex core (beginning from the first unsupported token, everything before that is still processed by the internal matcher). An example might make this more clear: |
|
454 |
- |
|
455 |
-*www\(\backslash\).google\(\backslash\).(com|ro|it) (\[a-zA-Z\])+\(\backslash\).google\(\backslash\).(com|ro|it)* |
|
456 |
- |
|
457 |
-Everything till *(\[a-zA-Z\])+* is processed internally, that parenthesis (and everything beyond) is processed by the posix core. |
|
458 |
- |
|
459 |
-Examples of url pairs that match: |
|
460 |
- |
|
461 |
-- *www.google.ro images.google.ro* |
|
462 |
- |
|
463 |
-- www.google.com images.google.ro |
|
464 |
- |
|
465 |
-Example of url pairs that don’t match: |
|
466 |
- |
|
467 |
-- www.google.ro images1.google.ro |
|
468 |
- |
|
469 |
-- images.google.com image.google.com |
|
470 |
- |
|
471 |
-## Flags |
|
472 |
- |
|
473 |
-Flags are a binary OR of the following numbers: |
|
474 |
- |
|
475 |
-- HOST_SUFFICIENT |
|
476 |
- |
|
477 |
- 1 |
|
478 |
- |
|
479 |
-- DOMAIN_SUFFICIENT |
|
480 |
- |
|
481 |
- 2 |
|
482 |
- |
|
483 |
-- DO_REVERSE_LOOKUP |
|
484 |
- |
|
485 |
- 4 |
|
486 |
- |
|
487 |
-- CHECK_REDIR |
|
488 |
- |
|
489 |
- 8 |
|
490 |
- |
|
491 |
-- CHECK_SSL |
|
492 |
- |
|
493 |
- 16 |
|
494 |
- |
|
495 |
-- CHECK_CLOAKING |
|
496 |
- |
|
497 |
- 32 |
|
498 |
- |
|
499 |
-- CLEANUP_URL |
|
500 |
- |
|
501 |
- 64 |
|
502 |
- |
|
503 |
-- CHECK_DOMAIN_REVERSE |
|
504 |
- |
|
505 |
- 128 |
|
506 |
- |
|
507 |
-- CHECK_IMG_URL |
|
508 |
- |
|
509 |
- 256 |
|
510 |
- |
|
511 |
-- DOMAINLIST_REQUIRED |
|
512 |
- |
|
513 |
- 512 |
|
514 |
- |
|
515 |
-The names of the constants are self-explanatory. |
|
516 |
- |
|
517 |
-These constants are defined in libclamav/phishcheck.h, you can check there for the latest flags. |
|
518 |
- |
|
519 |
-There is a default set of flags that are enabled, these are currently: |
|
520 |
- |
|
521 |
- ( CLEANUP_URL | CHECK_SSL | CHECK_CLOAKING | CHECK_IMG_URL ) |
|
522 |
- |
|
523 |
-ssl checking is performed only for a tags currently. |
|
524 |
- |
|
525 |
-You must decide for each line in the domainlist if you want to filter any flags (that is you don’t want certain checks to be done), and then calculate the binary OR of those constants, and then convert it into a 3-digit hexnumber. For example you devide that domain_sufficient shouldn’t be used for ebay.com, and you don’t want to check images either, so you come up with this flag number: \(2|256\Rightarrow\)258\((decimal)\Rightarrow102(hexadecimal)\) |
|
526 |
- |
|
527 |
-So you add this line to daily.wdb: |
|
528 |
- |
|
529 |
-- R102 www.ebay.com .+ |
|
530 |
- |
|
531 |
-# Introduction to regular expressions |
|
532 |
- |
|
533 |
-Recomended reading: |
|
534 |
- |
|
535 |
-- http://www.regular-expressions.info/quickstart.html |
|
536 |
- |
|
537 |
-- http://www.regular-expressions.info/tutorial.html |
|
538 |
- |
|
539 |
-- regex(7) man-page: http://www.tin.org/bin/man.cgi?section=7\&topic=regex |
|
540 |
- |
|
541 |
-## Special characters |
|
542 |
- |
|
543 |
-- \[ |
|
544 |
- |
|
545 |
- the opening square bracket - it marks the beginning of a character class, see section[Character-classes](#Character-classes) |
|
546 |
- |
|
547 |
-- \(\backslash\) |
|
548 |
- |
|
549 |
- the backslash - escapes special characters, see section [Escaping](#Escaping) |
|
550 |
- |
|
551 |
-- ^ |
|
552 |
- |
|
553 |
- the caret - matches the beginning of a line (not needed in clamav regexes, this is implied) |
|
554 |
- |
|
555 |
-- $ |
|
556 |
- |
|
557 |
- the dollar sign - matches the end of a line (not needed in clamav regexes, this is implied) |
|
558 |
- |
|
559 |
-- ̇ |
|
560 |
- |
|
561 |
- the period or dot - matches *any* character |
|
562 |
- |
|
563 |
-- | |
|
564 |
- |
|
565 |
- the vertical bar or pipe symbol - matches either of the token on its left and right side, see section [Alternation](#sub:Alternation) |
|
566 |
- |
|
567 |
-- ? |
|
568 |
- |
|
569 |
- the question mark - matches optionally the left-side token, see section[Optional-matching,-and](Optional-matching,-and) |
|
570 |
- |
|
571 |
-- \* |
|
572 |
- |
|
573 |
- the asterisk or star - matches 0 or more occurences of the left-side token, see section [Optional-matching,-and](Optional-matching,-and) |
|
574 |
- |
|
575 |
-- + |
|
576 |
- |
|
577 |
- the plus sign - matches 1 or more occurences of the left-side token, see section [Optional-matching,-and](Optional-matching,-and) |
|
578 |
- |
|
579 |
-- ( |
|
580 |
- |
|
581 |
- the opening round bracket - marks beginning of a group, see section [Groups](Groups) |
|
582 |
- |
|
583 |
-- ) |
|
584 |
- |
|
585 |
- the closing round bracket - marks end of a group, see section[Groups](Groups) |
|
586 |
- |
|
587 |
-## Character classes |
|
588 |
- |
|
589 |
-## Escaping |
|
590 |
- |
|
591 |
-Escaping has two purposes: |
|
592 |
- |
|
593 |
-- it allows you to actually match the special characters themselves, for example to match the literal *+*, you would write *\(\backslash\)+* |
|
594 |
- |
|
595 |
-- it also allows you to match non-printable characters, such as the tab (*\(\backslash\)t*), newline (*\(\backslash\)n*), .. |
|
596 |
- |
|
597 |
-However since non-printable characters are not valid inside an url, you won’t have a reason to use them. |
|
598 |
- |
|
599 |
-## Alternation |
|
600 |
- |
|
601 |
-## Optional matching, and repetition |
|
602 |
- |
|
603 |
-## Groups |
|
604 |
- |
|
605 |
-Groups are usually used together with repetition, or alternation. For example: *(com|it)+* means: match 1 or more repetitions of *com* or *it,* that is it matches: com, it, comcom, comcomcom, comit, itit, ititcom,... you get the idea. |
|
606 |
- |
|
607 |
-Groups can also be used to extract substring, but this is not supported by the clam engine, and not needed either in this case. |
|
608 |
- |
|
609 |
-# How to create database files |
|
610 |
- |
|
611 |
-## How to create and maintain the whitelist (daily.wdb) |
|
612 |
- |
|
613 |
-If the phishing code claims that a certain mail is phishing, but its not, you have 2 choices: |
|
614 |
- |
|
615 |
-- examine your rules daily.pdb, and fix them if necessary (see: section[How-to-create](How-to-create)) |
|
616 |
- |
|
617 |
-- add it to the whitelist (discussed here) |
|
618 |
- |
|
619 |
-Lets assume you are having problems because of links like this in a mail: |
|
620 |
- |
|
621 |
-```html |
|
622 |
- <a href=''http://69.0.241.57/bCentral/L.asp?L=XXXXXXXX''> |
|
623 |
- http://www.bcentral.it/ |
|
624 |
- </a> |
|
625 |
-``` |
|
626 |
- |
|
627 |
-After investigating those sites further, you decide they are no threat, and create a line like this in daily.wdb: |
|
628 |
- |
|
629 |
-``` |
|
630 |
-R http://www\(\backslash\).bcentral\(\backslash\).it/.+ |
|
631 |
-http://69\(\backslash\).0\(\backslash\).241\(\backslash\).57/bCentral/L\(\backslash\).asp?L=.+ |
|
632 |
-``` |
|
633 |
- |
|
634 |
-Note: urls like the above can be used to track unique mail recipients, and thus know if somebody actually reads mails (so they can send more spam). However since this site required no authentication information, it is safe from a phishing point of view. |
|
635 |
- |
|
636 |
-## How to create and maintain the domainlist (daily.pdb) |
|
637 |
- |
|
638 |
-When not using –phish-scan-alldomains (production environments for example), you need to decide which urls you are going to check. |
|
639 |
- |
|
640 |
-Although at a first glance it might seem a good idea to check everything, it would produce false positives. Particularly newsletters, ads, etc. are likely to use URLs that look like phishing attempts. |
|
641 |
- |
|
642 |
-Lets assume that you’ve recently seen many phishing attempts claiming they come from Paypal. Thus you need to add paypal to daily.pdb: |
|
643 |
- |
|
644 |
-``` |
|
645 |
-R .+ .+\(\backslash\).paypal\(\backslash\).com |
|
646 |
-``` |
|
647 |
- |
|
648 |
-The above line will block (detect as phishing) mails that contain urls that claim to lead to paypal, but they don’t in fact. |
|
649 |
- |
|
650 |
-Be carefull not to create regexes that match a too broad range of urls though. |
|
651 |
- |
|
652 |
-## Dealing with false positives, and undetected phishing mails |
|
653 |
- |
|
654 |
-### False positives |
|
655 |
- |
|
656 |
-Whenever you see a false positive (mail that is detected as phishing, but its not), you need to examine *why* clamav decided that its phishing. You can do this easily by building clamav with debugging (./configure –enable-experimental –enable-debug), and then running a tool: |
|
657 |
- |
|
658 |
-```bash |
|
659 |
-$contrib/phishing/why.py phishing.eml |
|
660 |
-``` |
|
661 |
- |
|
662 |
-This will show the url that triggers the phish verdict, and a reason why that url is considered phishing attempt. |
|
663 |
- |
|
664 |
-Once you know the reason, you might need to modify daily.pdb (if one of yours rules inthere are too broad), or you need to add the url to daily.wdb. If you think the algorithm is incorrect, please file a bug report on bugzilla.clamav.net, including the output of *why.py*. |
|
665 |
- |
|
666 |
-### Undetected phish mails |
|
667 |
- |
|
668 |
-Using why.py doesn’t help here unfortunately (it will say: clean), so all you can do is: |
|
669 |
- |
|
670 |
-```bash |
|
671 |
-$clamscan/clamscan –phish-scan-alldomains undetected.eml |
|
672 |
-``` |
|
673 |
- |
|
674 |
-And see if the mail is detected, if yes, then you need to add an appropriate line to daily.pdb (see section [How-to-create](How-to-create)). |
|
675 |
- |
|
676 |
-If the mail is not detected, then try using: |
|
677 |
- |
|
678 |
-```bash |
|
679 |
-$clamscan/clamscan –debug undetected.eml|less |
|
680 |
-``` |
|
681 |
- |
|
682 |
-Then see what urls are being checked, see if any of them is in a whitelist, see if all urls are detected, etc. |
... | ... |
@@ -3,849 +3,131 @@ |
3 | 3 |
Table of Contents |
4 | 4 |
|
5 | 5 |
- [Creating signatures for ClamAV](#creating-signatures-for-clamav) |
6 |
-- [Introduction](#introduction) |
|
7 |
-- [Signature formats](#signature-formats) |
|
8 |
- - [Hash-based signatures](#hash-based-signatures) |
|
9 |
- - [MD5 hash-based signatures](#md5-hash-based-signatures) |
|
10 |
- - [SHA1 and SHA256 hash-based signatures](#sha1-and-sha256-hash-based-signatures) |
|
11 |
- - [PE section based hash signatures](#pe-section-based-hash-signatures) |
|
12 |
- - [Hash signatures with unknown size](#hash-signatures-with-unknown-size) |
|
13 |
- - [Body-based signatures](#body-based-signatures) |
|
14 |
- - [Hexadecimal format](#hexadecimal-format) |
|
15 |
- - [Wildcards](#wildcards) |
|
16 |
- - [Character classes](#character-classes) |
|
17 |
- - [Alternate strings](#alternate-strings) |
|
18 |
- - [Basic signature format](#basic-signature-format) |
|
19 |
- - [Extended signature format](#extended-signature-format) |
|
20 |
- - [Logical signatures](#logical-signatures) |
|
21 |
- - [Subsignature Modifiers](#subsignature-modifiers) |
|
22 |
- - [Special Subsignature Types](#special-subsignature-types) |
|
23 |
- - [Macro subsignatures](#macro-subsignatures) |
|
24 |
- - [Byte Compare Subsignatures](#byte-compare-subsignatures) |
|
25 |
- - [PCRE subsignatures](#pcre-subsignatures) |
|
26 |
- - [Icon signatures for PE files](#icon-signatures-for-pe-files) |
|
27 |
- - [Signatures for Version Information metadata in PE files](#signatures-for-version-information-metadata-in-pe-files) |
|
28 |
- - [Trusted and Revoked Certificates](#trusted-and-revoked-certificates) |
|
29 |
- - [Signatures based on container metadata](#signatures-based-on-container-metadata) |
|
30 |
- - [Whitelist databases](#whitelist-databases) |
|
31 |
- - [Signature names](#signature-names) |
|
32 |
- - [Using YARA rules in ClamAV](#using-yara-rules-in-clamav) |
|
33 |
- - [Passwords for archive files \[experimental\]](#passwords-for-archive-files-experimental) |
|
34 |
-- [Signature writing tips and tricks](#signature-writing-tips-and-tricks) |
|
35 |
- - [Testing rules with clamscan](#testing-rules-with-clamscan) |
|
36 |
- - [Debug information from libclamav](#debug-information-from-libclamav) |
|
37 |
- - [Writing signatures for special files](#writing-signatures-for-special-files) |
|
38 |
- - [HTML](#html) |
|
39 |
- - [Text files](#text-files) |
|
40 |
- - [Compressed Portable Executable files](#compressed-portable-executable-files) |
|
41 |
- - [Using sigtool](#using-sigtool) |
|
42 |
- - [Inspecting signatures inside a CVD file](#inspecting-signatures-inside-a-CVD-file) |
|
43 |
- - [External tools](#external-tools) |
|
44 |
- |
|
45 |
-# Introduction |
|
6 |
+ - [Introduction](#introduction) |
|
7 |
+ - [Database formats](#database-formats) |
|
8 |
+ - [Settings databases](#settings-databases) |
|
9 |
+ - [Signature databases](#signature-databases) |
|
10 |
+ - [Body-based Signatures](#body-based-signatures) |
|
11 |
+ - [Hash-based Signatures](#hash-based-signatures) |
|
12 |
+ - [Alternative signature support](#alternative-signature-support) |
|
13 |
+ - [Other database files](#other-database-files) |
|
14 |
+ - [Signature names](#signature-names) |
|
15 |
+ - [Signature Writing Tips and Tricks](#signature-writing-tips-and-tricks) |
|
16 |
+ - [Testing rules with `clamscan`](#testing-rules-with-clamscan) |
|
17 |
+ - [Debug information from libclamav](#debug-information-from-libclamav) |
|
18 |
+ - [Writing signatures for special files](#writing-signatures-for-special-files) |
|
19 |
+ - [HTML](#html) |
|
20 |
+ - [Text files](#text-files) |
|
21 |
+ - [Compressed Portable Executable files](#compressed-portable-executable-files) |
|
22 |
+ - [Using `sigtool`](#using-sigtool) |
|
23 |
+ - [Inspecting signatures inside a CVD file](#inspecting-signatures-inside-a-cvd-file) |
|
24 |
+ - [External tools](#external-tools) |
|
25 |
+ |
|
26 |
+## Introduction |
|
46 | 27 |
|
47 | 28 |
In order to detect malware and other file-based threats, ClamAV relies on signatures to differentiate clean and malicious/unwanted files. ClamAV signatures are primarily text-based and conform to one of the ClamAV-specific signature formats associated with a given method of detection. These formats are explained in the [Signature formats](#signature-formats) section below. In addition, ClamAV 0.99 and above support signatures written in the YARA format. More information on this can be found in the [Using YARA rules in ClamAV](#using-yara-rules-in-clamav) section. |
48 | 29 |
|
49 | 30 |
The ClamAV project distributes a collection of signatures in the form of CVD (ClamAV Virus Database) files. The CVD file format provides a digitally-signed container that encapsulates the signatures and ensures that they can't be modified by a malicious third-party. This signature set is actively maintained by [Cisco Talos](https://www.talosintelligence.com/) and can be downloaded using the `freshclam` application that ships with ClamAV. For more details on this, see the [CVD file](#inspecting-signatures-inside-a-CVD-file) section. |
50 | 31 |
|
51 |
-# Signature formats |
|
32 |
+## Database formats |
|
52 | 33 |
|
53 |
-## Hash-based signatures |
|
34 |
+ClamAV CVD and CLD database archives may be unpacked to the current directory using `sigtool -u <database name>`. For more details on inspecting CVD and CLD files, see [Inspecting signatures inside a CVD file](#inspecting-signatures-inside-a-cvd-file). Once unpacked, you'll observe a large collection of database files with various extensions described below. |
|
54 | 35 |
|
55 |
-The easiest way to create signatures for ClamAV is to use filehash checksums, however this method can be only used against static malware. |
|
36 |
+The CVD and CLD database archives may be supplemented with custom database files in the formats described to gain additional detection functionality. This is done simply by adding files of the following formats to the database directory, typically `/usr/local/share/clamav` or `"C:\Program Files\ClamAV\database"`. Alternatively, `clamd` and `clamscan` can be instructed to load the database from an alternative database file or database directory manually using the `clamd` `DatabaseDirectory` config option or the `clamscan -d` command line option. |
|
56 | 37 |
|
57 |
-### MD5 hash-based signatures |
|
38 |
+### Settings databases |
|
58 | 39 |
|
59 |
-To create a MD5 signature for `test.exe` use the `--md5` option of |
|
60 |
-sigtool: |
|
40 |
+ClamAV provides a handful of configuration related databases along side the signature definitions. |
|
61 | 41 |
|
62 |
-```bash |
|
63 |
-zolw@localhost:/tmp/test$ sigtool --md5 test.exe > test.hdb |
|
64 |
-zolw@localhost:/tmp/test$ cat test.hdb |
|
65 |
-48c4533230e1ae1c118c741c0db19dfb:17387:test.exe |
|
66 |
-``` |
|
67 |
- |
|
68 |
-That’s it\! The signature is ready for use: |
|
69 |
- |
|
70 |
-```bash |
|
71 |
-zolw@localhost:/tmp/test$ clamscan -d test.hdb test.exe |
|
72 |
-test.exe: test.exe FOUND |
|
73 |
- |
|
74 |
-Known viruses: 1 |
|
75 |
-Scanned directories: 0 |
|
76 |
-Engine version: 0.92.1 |
|
77 |
-Scanned files: 1 |
|
78 |
-Infected files: 1 |
|
79 |
-Data scanned: 0.02 MB |
|
80 |
-Time: 0.024 sec (0 m 0 s) |
|
81 |
-``` |
|
82 |
- |
|
83 |
-You can change the name (by default sigtool uses the name of the file) and place it inside a `*.hdb` file. A single database file can include any number of signatures. To get them automatically loaded each time clamscan/clamd starts just copy the database file(s) into the local virus database directory (eg. /usr/local/share/clamav). |
|
84 |
- |
|
85 |
-*The hash-based signatures shall not be used for text files, HTML and any other data that gets internally preprocessed before pattern matching. If you really want to use a hash signature in such a case, run clamscan with –debug and –leave-temps flags as described above and create a signature for a preprocessed file left in /tmp. Please keep in mind that a hash signature will stop matching as soon as a single byte changes in the target file.* |
|
86 |
- |
|
87 |
-### SHA1 and SHA256 hash-based signatures |
|
88 |
- |
|
89 |
-ClamAV 0.98 has also added support for SHA1 and SHA256 file checksums. The format is the same as for MD5 file checksum. It can differentiate between them based on the length of the hash string in the signature. For best backwards compatibility, these should be placed inside a `*.hsb` file. The format is: |
|
90 |
- |
|
91 |
-``` |
|
92 |
-HashString:FileSize:MalwareName |
|
93 |
-``` |
|
94 |
- |
|
95 |
-### PE section based hash signatures |
|
96 |
- |
|
97 |
-You can create a hash signature for a specific section in a PE file. Such signatures shall be stored inside `.mdb` files in the following format: |
|
98 |
- |
|
99 |
-``` |
|
100 |
-PESectionSize:PESectionHash:MalwareName |
|
101 |
-``` |
|
102 |
- |
|
103 |
-The easiest way to generate MD5 based section signatures is to extract target PE sections into separate files and then run sigtool with the option `--mdb` |
|
104 |
- |
|
105 |
-ClamAV 0.98 has also added support for SHA1 and SHA256 section based signatures. The format is the same as for MD5 PE section based signatures. It can differentiate between them based on the length of the hash string in the signature. For best backwards compatibility, these should be placed inside a `*.msb` file. |
|
106 |
- |
|
107 |
-### Hash signatures with unknown size |
|
108 |
- |
|
109 |
-ClamAV 0.98 has also added support for hash signatures where the size is not known but the hash is. It is much more performance-efficient to use signatures with specific sizes, so be cautious when using this feature. For these cases, the ’\*’ character can be used in the size field. To ensure proper backwards compatibility with older versions of ClamAV, these signatures must have a minimum functional level of 73 or higher. Signatures that use the wildcard size without this level set will be rejected as malformed. |
|
110 |
- |
|
111 |
-``` |
|
112 |
-Sample .hsb signature matching any size |
|
113 |
-HashString:*:MalwareName:73 |
|
114 |
- |
|
115 |
-Sample .msb signature matching any size |
|
116 |
-*:PESectionHash:MalwareName:73 |
|
117 |
-``` |
|
118 |
- |
|
119 |
-## Body-based signatures |
|
120 |
- |
|
121 |
-ClamAV stores all body-based signatures in a hexadecimal format. In this section by a hex-signature we mean a fragment of malware’s body converted into a hexadecimal string which can be additionally extended using various wildcards. |
|
122 |
- |
|
123 |
-### Hexadecimal format |
|
124 |
- |
|
125 |
-You can use `sigtool --hex-dump` to convert any data into a hex-string: |
|
126 |
- |
|
127 |
-```bash |
|
128 |
-zolw@localhost:/tmp/test$ sigtool --hex-dump |
|
129 |
-How do I look in hex? |
|
130 |
-486f7720646f2049206c6f6f6b20696e206865783f0a |
|
131 |
-``` |
|
132 |
- |
|
133 |
-### Wildcards |
|
134 |
- |
|
135 |
-ClamAV supports the following wildcards for hex-signatures: |
|
136 |
- |
|
137 |
-- `??` |
|
138 |
- |
|
139 |
- Match any byte. |
|
140 |
- |
|
141 |
-- `a?` |
|
142 |
- |
|
143 |
- Match a high nibble (the four high bits). |
|
144 |
- **IMPORTANT NOTE:** The nibble matching is only available in |
|
145 |
- libclamav with the functionality level 17 and higher therefore |
|
146 |
- please only use it with .ndb signatures followed by ":17" |
|
147 |
- (MinEngineFunctionalityLevel, see [3.2.7](#ndb)). |
|
148 |
- |
|
149 |
-- `?a` |
|
150 |
- |
|
151 |
- Match a low nibble (the four low bits). |
|
152 |
- |
|
153 |
-- `*` |
|
154 |
- |
|
155 |
- Match any number of bytes. |
|
156 |
- |
|
157 |
-- `{n}` |
|
158 |
- |
|
159 |
- Match \(n\) bytes. |
|
160 |
- |
|
161 |
-- `{-n}` |
|
162 |
- |
|
163 |
- Match \(n\) or less bytes. |
|
164 |
- |
|
165 |
-- `{n-}` |
|
166 |
- |
|
167 |
- Match \(n\) or more bytes. |
|
168 |
- |
|
169 |
-- `{n-m}` |
|
170 |
- |
|
171 |
- Match between \(n\) and \(m\) bytes (\(m > n\)). |
|
172 |
- |
|
173 |
-- `HEXSIG[x-y]aa` or `aa[x-y]HEXSIG` |
|
174 |
- |
|
175 |
- Match aa anchored to a hex-signature, see |
|
176 |
- <https://bugzilla.clamav.net/show_bug.cgi?id=776> for discussion and |
|
177 |
- examples. |
|
178 |
- |
|
179 |
-The range signatures `*` and `{}` virtually separate a hex-signature into two parts, eg. `aabbcc*bbaacc` is treated as two sub-signatures `aabbcc` and `bbaacc` with any number of bytes between them. It’s a requirement that each sub-signature includes a block of two static characters somewhere in its body. Note that there is one exception to this restriction; that is when the range wildcard is of the form `{n}` with `n<128`. In this case, ClamAV uses an optimization and translates `{n}` to the string consisting of `n ??` character wildcards. Character wildcards do not divide hex signatures into two parts and so the two static character requirement does not apply. |
|
180 |
- |
|
181 |
-### Character classes |
|
182 |
- |
|
183 |
-ClamAV supports the following character classes for hex-signatures: |
|
184 |
- |
|
185 |
-- `(B)` |
|
186 |
- |
|
187 |
- Match word boundary (including file boundaries). |
|
188 |
- |
|
189 |
-- `(L)` |
|
190 |
- |
|
191 |
- Match CR, CRLF or file boundaries. |
|
192 |
- |
|
193 |
-- `(W)` |
|
194 |
- |
|
195 |
- Match a non-alphanumeric character. |
|
196 |
- |
|
197 |
-### Alternate strings |
|
198 |
- |
|
199 |
-- Single-byte alternates (clamav-0.96) `(aa|bb|cc|...)` or `!(aa|bb|cc|...)` Match a member from a set of bytes \[aa, bb, cc, ...\]. |
|
200 |
- - Negation operation can be applied to match any non-member, assumed to be one-byte in length. |
|
201 |
- - Signature modifiers and wildcards cannot be applied. |
|
202 |
- |
|
203 |
-- Multi-byte fixed length alternates `(aaaa|bbbb|cccc|...)` or `!(aaaa|bbbb|cccc|...)` Match a member from a set of multi-byte alternates \[aaaa, bbbb, cccc, ...\] of n-length. |
|
204 |
- - All set members must be the same length. |
|
205 |
- - Negation operation can be applied to match any non-member, assumed to be n-bytes in length (clamav-0.98.2). |
|
206 |
- - Signature modifiers and wildcards cannot be applied. |
|
207 |
- |
|
208 |
-- Generic alternates (clamav-0.99) `(alt1|alt2|alt3|...)` Match a member from a set of alternates \[alt1, alt2, alt3, ...\] that can be of variable lengths. |
|
209 |
- - Negation operation cannot be applied. |
|
210 |
- - Signature modifiers and nibble wildcards \[`??, a?, ?a`\] can be applied. |
|
211 |
- - Ranged wildcards \[`{n-m}`\] are limited to a fixed range of less than 128 bytes \[`{1} -> {127}`\]. |
|
212 |
- |
|
213 |
-Note that using signature modifiers and wildcards classifies the alternate type to be a generic alternate. Thus single-byte alternates and multi-byte fixed length alternates can use signature modifiers and wildcards but will be classified as generic alternate. This means that negation cannot be applied in this situation and there is a slight performance impact. |
|
214 |
- |
|
215 |
-### Basic signature format |
|
216 |
- |
|
217 |
-The simplest (and now deprecated) signature format is: |
|
218 |
- |
|
219 |
-``` |
|
220 |
-MalwareName=HexSignature |
|
221 |
-``` |
|
222 |
- |
|
223 |
-ClamAV will scan the entire file looking for HexSignature. All signatures of this type must be placed inside `*.db` files. |
|
224 |
- |
|
225 |
-### Extended signature format |
|
226 |
- |
|
227 |
-The extended signature format allows for specification of additional information such as a target file type, virus offset or engine version, making the detection more reliable. The format is: |
|
228 |
- |
|
229 |
-``` |
|
230 |
-MalwareName:TargetType:Offset:HexSignature[:MinFL:[MaxFL]] |
|
231 |
-``` |
|
232 |
- |
|
233 |
-where `TargetType` is one of the following numbers specifying the type of the target file: |
|
234 |
- |
|
235 |
-- 0 = any file |
|
236 |
- |
|
237 |
-- 1 = Portable Executable, both 32- and 64-bit. |
|
238 |
- |
|
239 |
-- 2 = OLE2 containers, including their specific macros. The OLE2 format is primarily used by MS Office and MSI installation files. |
|
240 |
- |
|
241 |
-- 3 = HTML (normalized: whitespace transformed to spaces, tags/tag attributes normalized, all lowercase), Javascript is normalized too: all strings are normalized (hex encoding is decoded), numbers are parsed and normalized, local variables/function names are normalized to ’n001’ format, argument to eval() is parsed as JS again, unescape() is handled, some simple JS packers are handled, output is whitespace normalized. |
|
242 |
- |
|
243 |
-- 4 = Mail file |
|
244 |
- |
|
245 |
-- 5 = Graphics |
|
246 |
- |
|
247 |
-- 6 = ELF |
|
248 |
- |
|
249 |
-- 7 = ASCII text file (normalized) |
|
250 |
- |
|
251 |
-- 8 = Unused |
|
252 |
- |
|
253 |
-- 9 = Mach-O files |
|
254 |
- |
|
255 |
-- 10 = PDF files |
|
256 |
- |
|
257 |
-- 11 = Flash files |
|
258 |
- |
|
259 |
-- 12 = Java class files |
|
260 |
- |
|
261 |
-And `Offset` is an asterisk or a decimal number `n` possibly combined with a special modifier: |
|
262 |
- |
|
263 |
-- `*` = any |
|
264 |
- |
|
265 |
-- `n` = absolute offset |
|
266 |
- |
|
267 |
-- `EOF-n` = end of file minus `n` bytes |
|
268 |
- |
|
269 |
-Signatures for PE, ELF and Mach-O files additionally support: |
|
270 |
- |
|
271 |
-- `EP+n` = entry point plus n bytes (`EP+0` for `EP`) |
|
272 |
- |
|
273 |
-- `EP-n` = entry point minus n bytes |
|
274 |
- |
|
275 |
-- `Sx+n` = start of section `x`’s (counted from 0) data plus `n` bytes |
|
276 |
- |
|
277 |
-- `SEx` = entire section `x` (offset must lie within section boundaries) |
|
278 |
- |
|
279 |
-- `SL+n` = start of last section plus `n` bytes |
|
280 |
- |
|
281 |
-All the above offsets except `*` can be turned into **floating offsets** and represented as `Offset,MaxShift` where `MaxShift` is an unsigned integer. A floating offset will match every offset between `Offset` and `Offset+MaxShift`, eg. `10,5` will match all offsets from 10 to 15 and `EP+n,y` will match all offsets from `EP+n` to `EP+n+y`. Versions of ClamAV older than 0.91 will silently ignore the `MaxShift` extension and only use `Offset`. Optional `MinFL` and `MaxFL` parameters can restrict the signature to specific engine releases. All signatures in the extended format must be placed inside `*.ndb` files. |
|
282 |
- |
|
283 |
-### Logical signatures |
|
284 |
- |
|
285 |
-Logical signatures allow combining of multiple signatures in extended format using logical operators. They can provide both more detailed and flexible pattern matching. The logical sigs are stored inside `*.ldb` files in the following format: |
|
286 |
- |
|
287 |
-``` |
|
288 |
-SignatureName;TargetDescriptionBlock;LogicalExpression;Subsig0; |
|
289 |
-Subsig1;Subsig2;... |
|
290 |
-``` |
|
291 |
- |
|
292 |
-where: |
|
293 |
- |
|
294 |
-- `TargetDescriptionBlock` provides information about the engine and target file with comma separated `Arg:Val` pairs. For args where `Val` is a range, the minimum and maximum values should be expressed as `min-max`. |
|
295 |
- |
|
296 |
-- `LogicalExpression` specifies the logical expression describing the relationship between `Subsig0...SubsigN`. **Basis clause:** 0,1,...,N decimal indexes are SUB-EXPRESSIONS representing `Subsig0, Subsig1,...,SubsigN` respectively. **Inductive clause:** if `A` and `B` are SUB-EXPRESSIONS and `X, Y` are decimal numbers then `(A&B)`, `(A|B)`, `A=X`, `A=X,Y`, `A>X`, `A>X,Y`, `A<X` and `A<X,Y` are SUB-EXPRESSIONS |
|
297 |
- |
|
298 |
-- `SubsigN` is n-th subsignature in extended format possibly preceded with an offset. There can be specified up to 64 subsigs. |
|
299 |
- |
|
300 |
-Keywords used in `TargetDescriptionBlock`: |
|
301 |
- |
|
302 |
-- `Target:X`: Target file type |
|
303 |
- |
|
304 |
-- `Engine:X-Y`: Required engine functionality (range; 0.96). Note that if the `Engine` keyword is used, it must be the first one in the `TargetDescriptionBlock` for backwards compatibility |
|
305 |
- |
|
306 |
-- `FileSize:X-Y`: Required file size (range in bytes; 0.96) |
|
307 |
- |
|
308 |
-- `EntryPoint`: Entry point offset (range in bytes; 0.96) |
|
309 |
- |
|
310 |
-- `NumberOfSections`: Required number of sections in executable (range; 0.96) |
|
311 |
- |
|
312 |
-- `Container:CL_TYPE_*`: File type of the container which stores the scanned file. |
|
313 |
- |
|
314 |
- Specifying `CL_TYPE_ANY` matches on root objects only (i.e. the target file is explicitely _not_ in a container). Chances slim that you would want to use `CL_TYPE_ANY` in a signature, because placing the malicious file in an archive will then prevent it from alerting. |
|
315 |
- |
|
316 |
- Every ClamAV file type has the potential to be a container for additional files, although some are more likely than others. When a file is parsed and data in the file is identified to be scanned as a unique type, that parent file becomes a container the moment the embedded content is scanned. For a list of possible CL_TYPEs, refer to the [File Types Reference](ClamAV-File-Types.md). |
|
317 |
- |
|
318 |
-- `Intermediates:CL_TYPE_*>CL_TYPE_*`: Specify one or more layers of file types containing the scanned file. _This is an alternative to using `Container`._ |
|
319 |
- |
|
320 |
- You may specify up to 16 layers of file types separated by ’`>`’ in top-down order. Note that the ’`>`’ separator is not needed if you only specify a single container. The last type should be the immediate container containing the malicious file. Unlike with the `Container` option, `CL_TYPE_ANY` can be used as a wildcard file type. (expr; 0.100.0) |
|
321 |
- |
|
322 |
- For a list of possible CL_TYPEs, refer to the [File Types Reference](ClamAV-File-Types.md). |
|
323 |
- |
|
324 |
-- `IconGroup1`: Icon group name 1 from .idb signature Required engine functionality (range; 0.96) |
|
325 |
- |
|
326 |
-- `IconGroup2`: Icon group name 2 from .idb signature Required engine functionality (range; 0.96) |
|
327 |
- |
|
328 |
-Modifiers for subexpressions: |
|
329 |
- |
|
330 |
-- `A=X`: If the SUB-EXPRESSION A refers to a single signature then this signature must get matched exactly X times; if it refers to a (logical) block of signatures then this block must generate exactly X matches (with any of its sigs). |
|
331 |
- |
|
332 |
-- `A=0` specifies negation (signature or block of signatures cannot be matched) |
|
333 |
- |
|
334 |
-- `A=X,Y`: If the SUB-EXPRESSION A refers to a single signature then this signature must be matched exactly X times; if it refers to a (logical) block of signatures then this block must generate X matches and at least Y different signatures must get matched. |
|
335 |
- |
|
336 |
-- `A>X`: If the SUB-EXPRESSION A refers to a single signature then this signature must get matched more than X times; if it refers to a (logical) block of signatures then this block must generate more than X matches (with any of its sigs). |
|
337 |
- |
|
338 |
-- `A>X,Y`: If the SUB-EXPRESSION A refers to a single signature then this signature must get matched more than X times; if it refers to a (logical) block of signatures then this block must generate more than X matches _and_ at least Y different signatures must be matched. |
|
339 |
- |
|
340 |
-- `A<X`: Just like `A>Z` above with the change of "more" to "less". |
|
341 |
- |
|
342 |
- If the SUB-EXPRESSION A refers to a single signature then this signature must get matched less than X times; if it refers to a (logical) block of signatures then this block must generate less than X matches (with any of its sigs). |
|
343 |
- |
|
344 |
-- `A<X,Y`: Similar to `A>X,Y`. If the SUB-EXPRESSION A refers to a single signature then this signature must get matched less than X times; if it refers to a (logical) block of signatures then this block must generate less than X matches _and_ at least Y different signatures must be matched. |
|
345 |
- |
|
346 |
-Examples: |
|
347 |
- |
|
348 |
-``` |
|
349 |
-Sig1;Target:0;(0&1&2&3)&(4|1);6b6f74656b;616c61;7a6f6c77;7374656 |
|
350 |
-6616e;deadbeef |
|
351 |
- |
|
352 |
-Sig2;Target:0;((0|1|2)>5,2)&(3|1);6b6f74656b;616c61;7a6f6c77;737 |
|
353 |
-46566616e |
|
354 |
- |
|
355 |
-Sig3;Target:0;((0|1|2|3)=2)&(4|1);6b6f74656b;616c61;7a6f6c77;737 |
|
356 |
-46566616e;deadbeef |
|
357 |
- |
|
358 |
-Sig4;Engine:51-255,Target:1;((0|1)&(2|3))&4;EP+123:33c06834f04100 |
|
359 |
-f2aef7d14951684cf04100e8110a00;S2+78:22??232c2d252229{-15}6e6573 |
|
360 |
-(63|64)61706528;S3+50:68efa311c3b9963cb1ee8e586d32aeb9043e;f9c58 |
|
361 |
-dcf43987e4f519d629b103375;SL+550:6300680065005c0046006900 |
|
362 |
-``` |
|
363 |
- |
|
364 |
-### Subsignature Modifiers |
|
365 |
- |
|
366 |
-ClamAV (clamav-0.99) supports a number of additional subsignature |
|
367 |
-modifiers for logical signatures. This is done by specifying `::` |
|
368 |
-followed by a number of characters representing the desired options. |
|
369 |
-Signatures using subsignature modifiers require `Engine:81-255` for |
|
370 |
-backwards-compatibility. |
|
371 |
- |
|
372 |
-- Case-Insensitive \[`i`\] |
|
373 |
- |
|
374 |
- Specifying the `i` modifier causes ClamAV to match all alphabetic hex bytes as case-insensitive. All patterns in ClamAV are case-sensitive by default. |
|
375 |
- |
|
376 |
-- Wide \[`w`\] |
|
377 |
- |
|
378 |
- Specifying the `w` causes ClamAV to match all hex bytes encoded with two bytes per character. Note this simply interweaves each character with NULL characters and does not truly support UTF-16 characters. Wildcards for ’wide’ subsignatures are not treated as wide (i.e. there can be an odd number of intermittent characters). This can be combined with `a` to search for patterns in both wide and ascii. |
|
379 |
- |
|
380 |
-- Fullword \[`f`\] |
|
381 |
- |
|
382 |
- Match subsignature as a fullword (delimited by non-alphanumeric characters). |
|
383 |
- |
|
384 |
-- Ascii \[`a`\] |
|
385 |
- |
|
386 |
- Match subsignature as ascii characters. This can be combined with `w` to search for patterns in both ascii and wide. |
|
387 |
- |
|
388 |
-Examples: |
|
389 |
- |
|
390 |
-``` |
|
391 |
-clamav-nocase-A;Engine:81-255,Target:0;0&1;41414141::i;424242424242::i |
|
392 |
- -matches 'AAAA'(nocase) and 'BBBBBB'(nocase) |
|
393 |
- |
|
394 |
-clamav-fullword-A;Engine:81-255,Target:0;0&1;414141;68656c6c6f::f |
|
395 |
- -matches 'AAA' and 'hello'(fullword) |
|
396 |
-clamav-fullword-B;Engine:81-255,Target:0;0&1;414141;68656c6c6f::fi |
|
397 |
- -matches 'AAA' and 'hello'(fullword nocase) |
|
398 |
- |
|
399 |
-clamav-wide-B2;Engine:81-255,Target:0;0&1;414141;68656c6c6f::wa |
|
400 |
- -matches 'AAA' and 'hello'(wide ascii) |
|
401 |
-clamav-wide-C0;Engine:81-255,Target:0;0&1;414141;68656c6c6f::iwfa |
|
402 |
- -matches 'AAA' and 'hello'(nocase wide fullword ascii) |
|
403 |
-``` |
|
404 |
- |
|
405 |
-## Special Subsignature Types |
|
406 |
- |
|
407 |
-### Macro subsignatures |
|
408 |
- |
|
409 |
-Introduced in ClamAV 0.96 |
|
410 |
- |
|
411 |
-Format: `${min-max}MACROID$` |
|
412 |
- |
|
413 |
-Macro subsignatures are used to combine a number of existing extended |
|
414 |
-signatures (`.ndb`) into a on-the-fly generated alternate string logical |
|
415 |
-signature (`.ldb`). Signatures using macro subsignatures require |
|
416 |
-`Engine:51-255` for backwards-compatibility. |
|
42 |
+`*.cfg`: [Dynamic config settings](Signatures/DynamicConfig.md) |
|
417 | 43 |
|
418 |
-Example: |
|
44 |
+`*.cat` `*.crb`: [Trusted and revoked PE certs](Signatures/AuthenticodeRules.md) |
|
419 | 45 |
|
420 |
-``` |
|
421 |
- test.ldb: |
|
422 |
- TestMacro;Engine:51-255,Target:0;0&1;616161;${6-7}12$ |
|
423 |
- |
|
424 |
- test.ndb: |
|
425 |
- D1:0:$12:626262 |
|
426 |
- D2:0:$12:636363 |
|
427 |
- D3:0:$30:626264 |
|
428 |
-``` |
|
429 |
- |
|
430 |
-The example logical signature `TestMacro` is functionally equivalent |
|
431 |
-to: |
|
432 |
- |
|
433 |
-``` |
|
434 |
-`TestMacro;Engine:51-255,Target:0;0;616161{3-4}(626262|636363)` |
|
435 |
-``` |
|
436 |
- |
|
437 |
-- `MACROID` points to a group of signatures; there can be at most 32 macro groups. |
|
438 |
- |
|
439 |
- - In the example, `MACROID` is `12` and both `D1` and `D2` are members of macro group `12`. `D3` is a member of separate macro group `30`. |
|
440 |
- |
|
441 |
-- `{min-max}` specifies the offset range at which one of the group signatures should match; the offset range is relative to the starting offset of the preceding subsignature. This means a macro subsignature cannot be the first subsignature. |
|
442 |
- |
|
443 |
- - In the example, `{min-max}` is `{6-7}` and it is relative to the start of a `616161` match. |
|
444 |
- |
|
445 |
-- For more information and examples please see <https://bugzilla.clamav.net/show_bug.cgi?id=164>. |
|
446 |
- |
|
447 |
-### Byte Compare Subsignatures |
|
448 |
- |
|
449 |
-Introduced in ClamAV 0.101 |
|
450 |
- |
|
451 |
-Format: `subsigid_trigger(offset#byte_options#comparisons)` |
|
452 |
- |
|
453 |
-Byte compare subsignatures can be used to evaluate a numeric value at a given offset from the start of another (matched) subsignature within the same logical signature. These are executed after all other subsignatures within the logical subsignature are fired, with the exception of PCRE subsignatures. They can evaluate offsets only from a single referenced subsignature, and that subsignature must give a valid match for the evaluation to occur. |
|
454 |
- |
|
455 |
-- `subsigid_trigger` is a required field and may refer to any single non-PCRE, non-Byte Compare subsignature within the lsig. The byte compare subsig will evaluate if `subsigid_trigger` matches. Triggering on multiple subsigs or logic based triggering is not currently supported. |
|
456 |
- |
|
457 |
-- `offset` is a required field that consists of an `offset_modifier` and a numeric `offset` (hex or decimal offsets are okay). |
|
458 |
- |
|
459 |
- - `offset_modifier` can be either `>>` or `<<` where the former denotes a positive offset and the latter denotes a negative offset. The offset is calculated from the start of `subsigid_trigger`, which allows for byte extraction before the specified match, after the match, and within the match itself. |
|
460 |
- |
|
461 |
- - `offset` must be a positive hex or decimal value. This will be the number of bytes from the start of the referenced `subsigid_trigger` match within the file buffer to begin the comparison. |
|
462 |
- |
|
463 |
-- `byte_options` are used to specify the numeric type and endianess of the extracted byte sequence in that order as well as the number of bytes to be read. By default ClamAV will attempt to matchup up to the number of byte specified, unless the `e` (exact) option is specified or the numeric type is `b` (binary). This field follows the form `[h|d|a|i][l|b][e]num_bytes` |
|
464 |
- |
|
465 |
- - `h|d|a|i` where `h` specifies the byte sequence will be in hex, `d` decimal, `a` automatic detection of hex or decimal at runtime, and `i` signifies raw binary data. |
|
466 |
- |
|
467 |
- - `l|b` where `l` specifies the byte sequence will be in little endian order and `b` big endian. If decimal `d` is specified, big-endian is implied and using `l` will result in a malformed database error. |
|
468 |
- |
|
469 |
- - `e` specifies that ClamAV will only evaluate the comparison if it can extract the exact number of bytes specified. This option is implicitly declared when using the `i` flag. |
|
470 |
- |
|
471 |
- - `num_bytes` specifies the number of bytes to extract. This can be a hex or decimal value. If `i` is specified only 1, 2, 4, and 8 are valid options. |
|
472 |
- |
|
473 |
-- `comparisons` are a required field which denotes how to evaluate the extracted byte sequence. Each Byte Compare signature can have one or two `comparison_sets` separated by a comma. Each `comparison_set` consists of a `Comparison_symbol` and a `Comparison_value` and takes the form `Comparison_symbolComparison_value`. Thus, `comparisons` takes the form `comparison_set[,comparison_set]` |
|
474 |
- |
|
475 |
- - `Comparison_symbol` denotes the type of comparison to be done. The supported comparison symbols are `<`, `>`, `=`. |
|
476 |
- |
|
477 |
- - `Comparison_value` is a required field which must be a numeric hex or decimal value. If all other conditions are met, the byte compare subsig will evalutate the extracted byte sequence against this number based on the provided `comparison_symbol`. |
|
478 |
- |
|
479 |
-### PCRE subsignatures |
|
480 |
- |
|
481 |
-Introduced in ClamAV 0.99 |
|
482 |
- |
|
483 |
-Format: `Trigger/PCRE/[Flags]` |
|
484 |
- |
|
485 |
-PCRE subsignatures are used within a logical signature (`.ldb`) to specify regex matches that execute once triggered by a conditional based on preceding subsignatures. Signatures using PCRE subsignatures require `Engine:81-255` for backwards-compatibility. |
|
486 |
- |
|
487 |
-- `Trigger` is a required field that is a valid `LogicalExpression` and may refer to any subsignatures that precede this subsignature. Triggers cannot be self-referential and cannot refer to subsequent subsignatures. |
|
488 |
- |
|
489 |
-- `PCRE` is the expression representing the regex to execute. `PCRE` must be delimited by ’/’ and usage of ’/’ within the expression need to be escaped. For backward compatibility, ’;’ within the expression must be expressed as ’`\x3B`’. `PCRE` cannot be empty and (?UTF\*) control sequence is not allowed. If debug is specified, named capture groups are displayed in a post-execution report. |
|
490 |
- |
|
491 |
-- `Flags` are a series of characters which affect the compilation and execution of `PCRE` within the PCRE compiler and the ClamAV engine. This field is optional. |
|
492 |
- |
|
493 |
- - `g [CLAMAV_GLOBAL]` specifies to search for ALL matches of PCRE (default is to search for first match). NOTE: INCREASES the time needed to run the PCRE. |
|
494 |
- |
|
495 |
- - `r [CLAMAV_ROLLING]` specifies to use the given offset as the starting location to search for a match as opposed to the only location; applies to subsigs without maxshifts. By default, in order to facilatate normal ClamAV offset behavior, PCREs are auto-anchored (only attempt match on first offset); using the rolling option disables the auto-anchoring. |
|
496 |
- |
|
497 |
- - `e [CLAMAV_ENCOMPASS]` specifies to CONFINE matching between the specified offset and maxshift; applies only when maxshift is specified. Note: DECREASES time needed to run the PCRE. |
|
498 |
- |
|
499 |
- - `i [PCRE_CASELESS]` |
|
500 |
- |
|
501 |
- - `s [PCRE_DOTALL]` |
|
502 |
- |
|
503 |
- - `m [PCRE_MULTILINE]` |
|
504 |
- |
|
505 |
- - `x [PCRE_EXTENDED]` |
|
46 |
+`*.ftm`: [File Type Magic (FTM)](Signatures/FileTypeMagic.md) |
|
506 | 47 |
|
507 |
- - `A [PCRE_ANCHORED]` |
|
48 |
+### Signature databases |
|
508 | 49 |
|
509 |
- - `E [PCRE_DOLLAR_ENODNLY]` |
|
50 |
+_Note_: Signature databases with an extension ending in `u` are only loaded when Potentially Unwanted Application (PUA) signatures are enabled (default: off). |
|
510 | 51 |
|
511 |
- - `U [PCRE_UNGREEDY]` |
|
52 |
+#### Body-based Signatures |
|
512 | 53 |
|
513 |
-Examples: |
|
54 |
+Body-based signature content is a definition that matches not based on a hash but based on the specific sequences of bytes exhibited by the target file. |
|
514 | 55 |
|
515 |
-``` |
|
516 |
-Find.All.ClamAV;Engine:81-255,Target:0;1;6265676c6164697427736e6f7462797465636f6465;0/clamav/g |
|
517 |
- |
|
518 |
-Find.ClamAV.OnlyAt.299;Engine:81-255,Target:0;2;7374756c747a67657473;7063726572656765786c6f6c;299:0&1/clamav/ |
|
519 |
- |
|
520 |
-Find.ClamAV.StartAt.300;Engine:81-255,Target:0;3;616c61696e;62756731393238;636c6f736564;300:0&1&2/clamav/r |
|
521 |
- |
|
522 |
-Find.All.Encompassed.ClamAV;Engine:81-255,Target:0;3;7768796172656e2774;796f757573696e67;79617261;200,300:0&1&2/clamav/ge |
|
523 |
- |
|
524 |
-Named.CapGroup.Pcre;Engine:81-255,Target:0;3;636f75727479617264;616c62756d;74657272696572;50:0&1&2/variable=(?<nilshell>.{16})end/gr |
|
525 |
- |
|
526 |
-Firefox.TreeRange.UseAfterFree;Engine:81-255,Target:0,Engine:81-255;0&1&2;2e766965772e73656c656374696f6e;2e696e76616c696461746553656c656374696f6e;0&1/\x2Eview\x2Eselection.*?\x2Etree\s*\x3D\s*null.*?\x2Einvalidate/smi |
|
527 |
- |
|
528 |
-Firefox.IDB.UseAfterFree;Engine:81-255,Target:0;0&1;4944424b657952616e6765;0/^\x2e(only|lowerBound|upperBound|bound)\x28.*?\x29.*?\x2e(lower|upper|lowerOpen|upperOpen)/smi |
|
529 |
- |
|
530 |
-Firefox.boundElements;Engine:81-255,Target:0;0&1&2;6576656e742e6 |
|
531 |
-26f756e64456c656d656e7473;77696e646f772e636c6f7365;0&1/on(load|click)\s*=\s*\x22?window\.close\s*\x28/si |
|
532 |
-``` |
|
533 |
- |
|
534 |
-## Icon signatures for PE files |
|
535 |
- |
|
536 |
-ClamAV 0.96 includes an approximate/fuzzy icon matcher to help detecting malicious executables disguising themselves as innocent looking image files, office documents and the like. |
|
537 |
- |
|
538 |
-Icon matching is only triggered via .ldb signatures using the special attribute tokens `IconGroup1` or `IconGroup2`. These identify two (optional) groups of icons defined in a .idb database file. The format of the .idb file is: |
|
539 |
- |
|
540 |
-``` |
|
541 |
-ICONNAME:GROUP1:GROUP2:ICON_HASH |
|
542 |
-``` |
|
543 |
- |
|
544 |
-where: |
|
545 |
- |
|
546 |
-- `ICON_NAME` is a unique string identifier for a specific icon, |
|
547 |
- |
|
548 |
-- `GROUP1` is a string identifier for the first group of icons (`IconGroup1`) |
|
549 |
- |
|
550 |
-- `GROUP2` is a string identifier for the second group of icons (`IconGroup2`), |
|
551 |
- |
|
552 |
-- `ICON_HASH` is a fuzzy hash of the icon image |
|
553 |
- |
|
554 |
-The `ICON_HASH` field can be obtained from the debug output of libclamav. For example: |
|
56 |
+ClamAV body-based signature content has a [special format](BodySignatureFormat.md) to allow regex-like matching of data that is not entirely known. This format is used extensively in both Extended Signatures and Logical Signatures. |
|
555 | 57 |
|
556 |
-```bash |
|
557 |
-LibClamAV debug: ICO SIGNATURE: |
|
558 |
-ICON_NAME:GROUP1:GROUP2:18e2e0304ce60a0cc3a09053a30000414100057e000afe0000e 80006e510078b0a08910d11ad04105e0811510f084e01040c080a1d0b0021000a39002a41 |
|
559 |
-``` |
|
58 |
+`*.ndb` `*.ndu`: [Extended signatures](Signatures/ExtendedSignatures.md) |
|
560 | 59 |
|
561 |
-## Signatures for Version Information metadata in PE files |
|
60 |
+`*.ldb` `*.ldu`; `*.idb`: [Logical Signatures](Signatures/LogicalSignatures.md) |
|
562 | 61 |
|
563 |
-Starting with ClamAV 0.96 it is possible to easily match certain information built into PE files (executables and dynamic link libraries). Whenever you lookup the properties of a PE executable file in windows, you are presented with a bunch of details about the file itself. |
|
62 |
+`*.cdb`: [Container Metadata Signatures](Signatures/ContainerMetadata.md) |
|
564 | 63 |
|
565 |
-These info are stored in a special area of the file resources which goes under the name of `VS_VERSION_INFORMATION` (or versioninfo for short). It is divided into 2 parts. The first part (which is rather uninteresting) is really a bunch of numbers and flags indicating the product and file version. It was originally intended for use with installers which, after parsing it, should be able to determine whether a certain executable or library are to be upgraded/overwritten or are already up to date. Suffice to say, this approach never really worked and is generally never used. |
|
64 |
+`*.cbc`: [Bytecode Signatures](Signatures/BytecodeSignatures.md) |
|
566 | 65 |
|
567 |
-The second block is much more interesting: it is a simple list of key/value strings, intended for user information and completely ignored by the OS. For example, if you look at ping.exe you can see the company being *"Microsoft Corporation"*, the description *"TCP/IP Ping command"*, the internal name *"ping.exe"* and so on... Depending on the OS version, some keys may be given peculiar visibility in the file properties dialog, however they are internally all the same. |
|
66 |
+`*.pdb` `*.gdb` `*.wdb`: [Phishing URL Signatures](Signatures/PhishSigs.md) |
|
568 | 67 |
|
569 |
-To match a versioninfo key/value pair, the special file offset anchor `VI` was introduced. This is similar to the other anchors (like `EP` and `SL`) except that, instead of matching the hex pattern against a single offset, it checks it against each and every key/value pair in the file. The `VI` token doesn’t need nor accept a `+/-` offset like e.g. `EP+1`. As for the hex signature itself, it’s just the utf16 dump of the key and value. Only the `??` and `(aa|bb)` wildcards are allowed in the signature. Usually, you don’t need to bother figuring it out: each key/value pair together with the corresponding VI-based signature is printed by `clamscan` when the `--debug` option is given. |
|
68 |
+#### Hash-based Signatures |
|
570 | 69 |
|
571 |
-For example `clamscan --debug freecell.exe` produces: |
|
70 |
+`*.hdb` `*.hsb` `*.hdu` `*.hsu`: File hash signatures |
|
572 | 71 |
|
573 |
-```bash |
|
574 |
-[...] |
|
575 |
-Recognized MS-EXE/DLL file |
|
576 |
-in cli_peheader |
|
577 |
-versioninfo_cb: type: 10, name: 1, lang: 410, rva: 9608 |
|
578 |
-cli_peheader: parsing version info @ rva 9608 (1/1) |
|
579 |
-VersionInfo (d2de): 'CompanyName'='Microsoft Corporation' - |
|
580 |
-VI:43006f006d00700061006e0079004e0061006d006500000000004d006900 |
|
581 |
-630072006f0073006f0066007400200043006f00720070006f0072006100740 |
|
582 |
-069006f006e000000 |
|
583 |
-VersionInfo (d32a): 'FileDescription'='Entertainment Pack |
|
584 |
-FreeCell Game' - VI:460069006c006500440065007300630072006900700 |
|
585 |
-0740069006f006e000000000045006e007400650072007400610069006e006d |
|
586 |
-0065006e00740020005000610063006b0020004600720065006500430065006 |
|
587 |
-c006c002000470061006d0065000000 |
|
588 |
-VersionInfo (d396): 'FileVersion'='5.1.2600.0 (xpclient.010817 |
|
589 |
--1148)' - VI:460069006c006500560065007200730069006f006e00000000 |
|
590 |
-0035002e0031002e0032003600300030002e003000200028007800700063006 |
|
591 |
-c00690065006e0074002e003000310030003800310037002d00310031003400 |
|
592 |
-380029000000 |
|
593 |
-VersionInfo (d3fa): 'InternalName'='freecell' - VI:49006e007400 |
|
594 |
-650072006e0061006c004e0061006d006500000066007200650065006300650 |
|
595 |
-06c006c000000 |
|
596 |
-VersionInfo (d4ba): 'OriginalFilename'='freecell' - VI:4f007200 |
|
597 |
-6900670069006e0061006c00460069006c0065006e0061006d0065000000660 |
|
598 |
-0720065006500630065006c006c000000 |
|
599 |
-VersionInfo (d4f6): 'ProductName'='Sistema operativo Microsoft |
|
600 |
-Windows' - VI:500072006f0064007500630074004e0061006d00650000000 |
|
601 |
-000530069007300740065006d00610020006f00700065007200610074006900 |
|
602 |
-76006f0020004d006900630072006f0073006f0066007400ae0020005700690 |
|
603 |
-06e0064006f0077007300ae000000 |
|
604 |
-VersionInfo (d562): 'ProductVersion'='5.1.2600.0' - VI:50007200 |
|
605 |
-6f006400750063007400560065007200730069006f006e00000035002e00310 |
|
606 |
-02e0032003600300030002e0030000000 |
|
607 |
-[...] |
|
608 |
-``` |
|
72 |
+`*.mdb` `*.msb` `*.mdu` `*.msu`: PE section hash signatures |
|
609 | 73 |
|
610 |
-Although VI-based signatures are intended for use in logical signatures you can test them using ordinary `.ndb` files. For example: |
|
74 |
+[Hash-based Signature format](Signatures/HashSignatures.md) |
|
611 | 75 |
|
612 |
-``` |
|
613 |
-my_test_vi_sig:1:VI:paste_your_hex_sig_here |
|
614 |
-``` |
|
615 |
- |
|
616 |
-Final note. If you want to decode a VI-based signature into a human readable form you can use: |
|
617 |
- |
|
618 |
-```bash |
|
619 |
-echo hex_string | xxd -r -p | strings -el |
|
620 |
-``` |
|
621 |
- |
|
622 |
-For example: |
|
623 |
- |
|
624 |
-```bash |
|
625 |
-$ echo 460069006c0065004400650073006300720069007000740069006f006e |
|
626 |
-000000000045006e007400650072007400610069006e006d0065006e007400200 |
|
627 |
-05000610063006b0020004600720065006500430065006c006c00200047006100 |
|
628 |
-6d0065000000 | xxd -r -p | strings -el |
|
629 |
-FileDescription |
|
630 |
-Entertainment Pack FreeCell Game |
|
631 |
-``` |
|
632 |
- |
|
633 |
-## Trusted and Revoked Certificates |
|
634 |
- |
|
635 |
-Clamav 0.98 checks signed PE files for certificates and verifies each certificate in the chain against a database of trusted and revoked certificates. The signature format is |
|
636 |
- |
|
637 |
-``` |
|
638 |
- Name;Trusted;Subject;Serial;Pubkey;Exponent;CodeSign;TimeSign;CertSign; |
|
639 |
- NotBefore;Comment[;minFL[;maxFL]] |
|
640 |
-``` |
|
76 |
+#### Alternative signature support |
|
641 | 77 |
|
642 |
-where the corresponding fields are: |
|
78 |
+`*.yar` `*.yara`: [Yara rules](Signatures/YaraRules.md) |
|
643 | 79 |
|
644 |
-- `Name:` name of the entry |
|
80 |
+### Other database files |
|
645 | 81 |
|
646 |
-- `Trusted:` bit field, specifying whether the cert is trusted. 1 for trusted. 0 for revoked |
|
82 |
+`*.fp` `*.sfp` `*.ign` `*.ign2`: [Whitelisted files, signatures](Signatures/Whitelists.md) |
|
647 | 83 |
|
648 |
-- `Subject:` sha1 of the Subject field in hex |
|
84 |
+`*.pwdb`: [Encrypted archive passwords](Signatures/EncryptedArchives.md) |
|
649 | 85 |
|
650 |
-- `Serial:` the serial number as clamscan –debug –verbose reports |
|
86 |
+`*.info`: [Database information](Signatures/DatabaseInfo.md)` |
|
651 | 87 |
|
652 |
-- `Pubkey:` the public key in hex |
|
653 |
- |
|
654 |
-- `Exponent:` the exponent in hex. Currently ignored and hardcoded to 010001 (in hex) |
|
655 |
- |
|
656 |
-- `CodeSign:` bit field, specifying whether this cert can sign code. 1 for true, 0 for false |
|
657 |
- |
|
658 |
-- `TimeSign:` bit field. 1 for true, 0 for false |
|
659 |
- |
|
660 |
-- `CertSign:` bit field, specifying whether this cert can sign other certs. 1 for true, 0 for false |
|
661 |
- |
|
662 |
-- `NotBefore:` integer, cert should not be added before this variable. Defaults to 0 if left empty |
|
663 |
- |
|
664 |
-- `Comment:` comments for this entry |
|
665 |
- |
|
666 |
-The signatures for certs are stored inside `.crb` files. |
|
667 |
- |
|
668 |
-## Signatures based on container metadata |
|
669 |
- |
|
670 |
-ClamAV 0.96 allows creating generic signatures matching files stored inside different container types which meet specific conditions. The signature format is |
|
671 |
- |
|
672 |
-``` |
|
673 |
- VirusName:ContainerType:ContainerSize:FileNameREGEX: |
|
674 |
- FileSizeInContainer:FileSizeReal:IsEncrypted:FilePos: |
|
675 |
- Res1:Res2[:MinFL[:MaxFL]] |
|
676 |
-``` |
|
677 |
- |
|
678 |
-where the corresponding fields are: |
|
679 |
- |
|
680 |
-- `VirusName:` Virus name to be displayed when signature matches |
|
681 |
- |
|
682 |
-- `ContainerType:` The file type containing the target file. For example: |
|
683 |
- - `CL_TYPE_ZIP`, |
|
684 |
- - `CL_TYPE_RAR`, |
|
685 |
- - `CL_TYPE_ARJ`, |
|
686 |
- - `CL_TYPE_MSCAB`, |
|
687 |
- - `CL_TYPE_7Z`, |
|
688 |
- - `CL_TYPE_MAIL`, |
|
689 |
- - `CL_TYPE_(POSIX|OLD)_TAR`, |
|
690 |
- - `CL_TYPE_CPIO_(OLD|ODC|NEWC|CRC)` |
|
691 |
- |
|
692 |
- Use `*` as a wild card to indicate that container type may be any file type. |
|
693 |
- For a full list of ClamAV file types, see the [ClamAV File Types Reference](ClamAV-File-Types.md) |
|
694 |
- |
|
695 |
-- `ContainerSize:` size of the container file itself (eg. size of the zip archive) specified in bytes as absolute value or range `x-y` |
|
696 |
- |
|
697 |
-- `FileNameREGEX:` regular expression describing name of the target file |
|
698 |
- |
|
699 |
-- `FileSizeInContainer:` usually compressed size; for MAIL, TAR and CPIO == `FileSizeReal`; specified in bytes as absolute value or range |
|
700 |
- |
|
701 |
-- `FileSizeReal:` usually uncompressed size; for MAIL, TAR and CPIO == `FileSizeInContainer`; absolute value or range |
|
702 |
- |
|
703 |
-- `IsEncrypted:` 1 if the target file is encrypted, 0 if it’s not and `*` to ignore |
|
704 |
- |
|
705 |
-- `FilePos:` file position in container (counting from 1); absolute value or range |
|
706 |
- |
|
707 |
-- `Res1:` when `ContainerType` is `CL_TYPE_ZIP` or `CL_TYPE_RAR` this field is treated as a CRC sum of the target file specified in hexadecimal format; for other container types it’s ignored |
|
708 |
- |
|
709 |
-- `Res2:` not used as of ClamAV 0.96 |
|
710 |
- |
|
711 |
-The signatures for container files are stored inside `.cdb` files. |
|
712 |
- |
|
713 |
-## Whitelist databases |
|
714 |
- |
|
715 |
-To whitelist a specific file use the MD5 signature format and place it inside a database file with the extension of `.fp`. To whitelist a specific file with the SHA1 or SHA256 file hash signature format, place the signature inside a database file with the extension of `.sfp`. To whitelist a specific signature from the database you just add its name into a local file called local.ign2 stored inside the database directory. You can additionally follow the signature name with the MD5 of the entire database entry for this signature, eg: |
|
716 |
- |
|
717 |
-``` |
|
718 |
- Eicar-Test-Signature:bc356bae4c42f19a3de16e333ba3569c |
|
719 |
-``` |
|
720 |
- |
|
721 |
-In such a case, the signature will no longer be whitelisted when its entry in the database gets modified (eg. the signature gets updated to avoid false alerts). |
|
722 |
- |
|
723 |
-## Signature names |
|
88 |
+### Signature names |
|
724 | 89 |
|
725 | 90 |
ClamAV uses the following prefixes for signature names: |
726 | 91 |
|
727 | 92 |
- *Worm* for Internet worms |
728 |
- |
|
729 | 93 |
- *Trojan* for backdoor programs |
730 |
- |
|
731 | 94 |
- *Adware* for adware |
732 |
- |
|
733 | 95 |
- *Flooder* for flooders |
734 |
- |
|
735 | 96 |
- *HTML* for HTML files |
736 |
- |
|
737 | 97 |
- *Email* for email messages |
738 |
- |
|
739 | 98 |
- *IRC* for IRC trojans |
740 |
- |
|
741 | 99 |
- *JS* for Java Script malware |
742 |
- |
|
743 | 100 |
- *PHP* for PHP malware |
744 |
- |
|
745 | 101 |
- *ASP* for ASP malware |
746 |
- |
|
747 | 102 |
- *VBS* for VBS malware |
748 |
- |
|
749 | 103 |
- *BAT* for BAT malware |
750 |
- |
|
751 | 104 |
- *W97M*, *W2000M* for Word macro viruses |
752 |
- |
|
753 | 105 |
- *X97M*, *X2000M* for Excel macro viruses |
754 |
- |
|
755 | 106 |
- *O97M*, *O2000M* for generic Office macro viruses |
756 |
- |
|
757 | 107 |
- *DoS* for Denial of Service attack software |
758 |
- |
|
759 | 108 |
- *DOS* for old DOS malware |
760 |
- |
|
761 | 109 |
- *Exploit* for popular exploits |
762 |
- |
|
763 | 110 |
- *VirTool* for virus construction kits |
764 |
- |
|
765 | 111 |
- *Dialer* for dialers |
766 |
- |
|
767 | 112 |
- *Joke* for hoaxes |
768 | 113 |
|
769 | 114 |
Important rules of the naming convention: |
770 | 115 |
|
771 | 116 |
- always use a -zippwd suffix in the malware name for signatures of type zmd, |
772 |
- |
|
773 | 117 |
- always use a -rarpwd suffix in the malware name for signatures of type rmd, |
774 |
- |
|
775 | 118 |
- only use alphanumeric characters, dash (-), dot (.), underscores (_) in malware names, never use space, apostrophe or quote mark. |
776 | 119 |
|
777 |
-## Using YARA rules in ClamAV |
|
778 |
- |
|
779 |
-ClamAV version 0.99 and above can process YARA rules. ClamAV virus database file names ending with “.yar” or “.yara” are parsed as yara rule files. The link to the YARA rule grammar documentation may be found at http://plusvic.github.io/yara/. There are currently a few limitations on using YARA rules within ClamAV: |
|
780 |
- |
|
781 |
-- YARA modules are not yet supported by ClamAV. This includes the “import” keyword and any YARA module-specific keywords. |
|
782 |
- |
|
783 |
-- Global rules(“global” keyword) are not supported by ClamAV. |
|
784 |
- |
|
785 |
-- External variables(“contains” and “matches” keywords) are not supported. |
|
120 |
+## Signature Writing Tips and Tricks |
|
786 | 121 |
|
787 |
-- YARA rules pre-compiled with the *yarac* command are not supported. |
|
122 |
+### Testing rules with `clamscan` |
|
788 | 123 |
|
789 |
-- As in the ClamAV logical and extended signature formats, YARA strings and segments of strings separated by wild cards must represent at least two octets of data. |
|
124 |
+To test a new signature, first create a text file with the extension corresponding to the signature type (Ex: `.ldb` for logical signatures). Then, add the signature as it's own line within the file. This file can be passed to `clamscan` via the `-d` option, which tells ClamAV to load signatures from the file specified. If the signature is not formatted correctly, ClamAV will display an error - run `clamscan` with `--debug --verbose` to see additional information about the error message. Some common causes of errors include: |
|
790 | 125 |
|
791 |
-- There is a maximum of 64 strings per YARA rule. |
|
126 |
+- The signature file has the incorrect extension type for the signatures contained within |
|
127 |
+- The file has one or more blank lines |
|
128 |
+- For logical signatures, a semicolon exists at the end of the file |
|
792 | 129 |
|
793 |
-- YARA rules in ClamAV must contain at least one literal, hexadecimal, or regular expression string. |
|
794 |
- |
|
795 |
-In addition, there are a few more ClamAV processing modes that may affect the outcome of YARA rules. |
|
796 |
- |
|
797 |
-- *File decomposition and decompression* - Since ClamAV uses file decomposition and decompression to find viruses within de-archived and uncompressed inner files, YARA rules executed by ClamAV will match against these files as well. |
|
798 |
- |
|
799 |
-- *Normalization* - By default, ClamAV normalizes HTML, JavaScript, and ASCII text files. YARA rules in ClamAV will match against the normalized result. The effects of normalization of these file types may be captured using `clamscan --leave-temps --tempdir=mytempdir`. YARA rules may then be written using the normalized file(s) found in `mytempdir`. Alternatively, starting with ClamAV 0.100.0, `clamscan --normalize=no` will prevent normalization and only scan the raw file. To obtain similar behavior prior to 0.99.2, use `clamscan --scan-html=no`. The corresponding parameters for clamd.conf are `Normalize` and `ScanHTML`. |
|
800 |
- |
|
801 |
-- *YARA conditions driven by string matches* - All YARA conditions are driven by string matches in ClamAV. This saves from executing every YARA rule on every file. Any YARA condition may be augmented with a string match clause which is always true, such as: |
|
802 |
- |
|
803 |
-```yara |
|
804 |
- rule CheckFileSize |
|
805 |
- { |
|
806 |
- strings: |
|
807 |
- $abc = "abc" |
|
808 |
- condition: |
|
809 |
- ($abc or not $abc) and filesize < 200KB |
|
810 |
- } |
|
811 |
-``` |
|
812 |
- |
|
813 |
-This will ensure that the YARA condition always performs the desired action (checking the file size in this example), |
|
814 |
- |
|
815 |
-## Passwords for archive files \[experimental\] |
|
816 |
- |
|
817 |
-ClamAV 0.99 allows for users to specify password attempts for certain password-compatible archives. Passwords will be attempted in order of appearance in the password signature file which use the extension of `.pwdb`. If no passwords apply or none are provided, ClamAV will default to the original behavior of parsing the file. Currently, as of ClamAV 0.99 \[flevel 81\], only `.zip` archives using the traditional PKWARE encryption are supported. The signature format is |
|
818 |
- |
|
819 |
-``` |
|
820 |
- SignatureName;TargetDescriptionBlock;PWStorageType;Password |
|
821 |
-``` |
|
822 |
- |
|
823 |
-where: |
|
824 |
- |
|
825 |
-- `SignatureName`: name to be displayed during debug when a password is successful |
|
826 |
- |
|
827 |
-- `TargetDescriptionBlock`: provides information about the engine and target file with comma separated Arg:Val pairs |
|
828 |
- - `Engine:X-Y`: Required engine functionality |
|
829 |
- - `Container:CL_TYPE_*`: File type of applicable containers |
|
830 |
- |
|
831 |
-- `PWStorageType`: determines how the password field is parsed |
|
832 |
- - 0 = cleartext |
|
833 |
- - 1 = hex |
|
834 |
- |
|
835 |
-- `Password`: value used in password attempt |
|
836 |
- |
|
837 |
-The signatures for password attempts are stored inside `.pwdb` files. |
|
838 |
- |
|
839 |
-# Signature writing tips and tricks |
|
840 |
-## Testing rules with clamscan |
|
841 |
- |
|
842 |
-To test a new signature, first create a text file with the extension corresponding to the signature type (Ex: '.lsb' for logical signatures). Then, add the signature as it's own line within the file. This file can be passed to `clamscan` via the `-d` option, which tells ClamAV to load signatures from the file specified. If the signature is not formatted correctly, ClamAV will display an error - run `clamscan` with `--debug --verbose` to see additional information about the error message. Some common causes of errors include: |
|
843 |
- - The signature file has the incorrect extension type for the signatures contained within |
|
844 |
- - The file has one or more blank lines |
|
845 |
- - For logical signatures, a semicolon exists at the end of the file |
|
846 |
- |
|
847 |
-If the rule is formatted correctly, clamscan will load the signature(s) in and scan any files specified via the command line invocation (or all files in the current directory if none are specified). A successful detection will look like the following: |
|
130 |
+If the rule is formatted correctly, `clamscan` will load the signature(s) in and scan any files specified via the command line invocation (or all files in the current directory if none are specified). A successful detection will look like the following: |
|
848 | 131 |
|
849 | 132 |
```bash |
850 | 133 |
clamscan -d test.ldb text.exe |
... | ... |
@@ -863,18 +145,21 @@ Time: 0.400 sec (0 m 0 s) |
863 | 863 |
``` |
864 | 864 |
|
865 | 865 |
If the rule did not match as intended: |
866 |
- - The file may have exceeded one or more of the default scanning limits built-in to ClamAV. Try running clamscan with the following options to see if raising the limits addresses the issue: `--max-filesize=2000M --max-scansize=2000M --max-files=2000000 --max-recursion=2000000 --max-embeddedpe=2000M --max-htmlnormalize=2000000 --max-htmlnotags=2000000 --max-scriptnormalize=2000000 --max-ziptypercg=2000000 --max-partitions=2000000 --max-iconspe=2000000 --max-rechwp3=2000000 --pcre-match-limit=2000000 --pcre-recmatch-limit=2000000 --pcre-max-filesize=2000M`. |
|
867 |
- - If matching on HTML or text files, ClamAV might be performing normalization that causes the content of the scanned file to change. See the [HTML](#html) and [Text file](#text-file) sections for more details. |
|
868 |
- - libclamav may have been unable to unpack or otherwise process the file. See [Debug information from libclamav](#debug-information-from-libclamav) for more details. |
|
866 |
+ |
|
867 |
+- The file may have exceeded one or more of the default scanning limits built-in to ClamAV. Try running `clamscan` with the following options to see if raising the limits addresses the issue: `--max-filesize=2000M --max-scansize=2000M --max-files=2000000 --max-recursion=2000000 --max-embeddedpe=2000M --max-htmlnormalize=2000000 --max-htmlnotags=2000000 --max-scriptnormalize=2000000 --max-ziptypercg=2000000 --max-partitions=2000000 --max-iconspe=2000000 --max-rechwp3=2000000 --pcre-match-limit=2000000 --pcre-recmatch-limit=2000000 --pcre-max-filesize=2000M`. |
|
868 |
+- If matching on HTML or text files, ClamAV might be performing normalization that causes the content of the scanned file to change. See the [HTML](#html) and [Text file](#text-file) sections for more details. |
|
869 |
+- libclamav may have been unable to unpack or otherwise process the file. See [Debug information from libclamav](#debug-information-from-libclamav) for more details. |
|
869 | 870 |
|
870 | 871 |
NOTE: If you run `clamscan` with a `-d` flag, ClamAV will not load in the signatures downloaded via `freshclam`. This means that: |
871 |
- - some of ClamAV's unpacking support might be disabled, since some unpackers are implemented as bytecode signatures |
|
872 |
- - PE whitelisting based on Authenticode signatures won't work, since this functionality relies on .crb rules |
|
872 |
+ |
|
873 |
+- some of ClamAV's unpacking support might be disabled, since some unpackers are implemented as bytecode signatures |
|
874 |
+- PE whitelisting based on Authenticode signatures won't work, since this functionality relies on `.crb` rules |
|
875 |
+ |
|
873 | 876 |
If any of this functionality is needed, load in the CVD files manually with additional `-d` flags. |
874 | 877 |
|
875 | 878 |
### Debug information from libclamav |
876 | 879 |
|
877 |
-In order to create efficient signatures for ClamAV it’s important to understand how the engine handles input files. The best way to see how it works is having a look at the debug information from libclamav. You can do it by calling `clamscan` with the `--debug` and `--leave-temps` flags. The first switch makes clamscan display all the interesting information from libclamav and the second one avoids deleting temporary files so they can be analyzed further. |
|
880 |
+In order to create efficient signatures for ClamAV it’s important to understand how the engine handles input files. The best way to see how it works is having a look at the debug information from libclamav. You can do it by calling `clamscan` with the `--debug` and `--leave-temps` flags. The first switch makes `clamscan` display all the interesting information from libclamav and the second one avoids deleting temporary files so they can be analyzed further. |
|
878 | 881 |
|
879 | 882 |
The now important part of the info is: |
880 | 883 |
|
... | ... |
@@ -1025,9 +310,9 @@ No additional files get created by libclamav. By writing a signature for the dec |
1025 | 1025 |
|
1026 | 1026 |
This method should be applied to all files for which you want to create signatures. By analyzing the debug information you can quickly see how the engine recognizes and preprocesses the data and what additional files get created. Signatures created for bottom-level temporary files are usually more generic and should help detecting the same malware in different forms. |
1027 | 1027 |
|
1028 |
-## Writing signatures for special files |
|
1028 |
+### Writing signatures for special files |
|
1029 | 1029 |
|
1030 |
-### HTML |
|
1030 |
+#### HTML |
|
1031 | 1031 |
|
1032 | 1032 |
ClamAV contains HTML normalization code which makes it easier to write signatures for HTML data that might differ based on white space, capitalization, and other insignificant differences. Running `sigtool --html-normalise` on a HTML file can be used to see what a file's contents will look like after normalization. This command should generate the following files: |
1033 | 1033 |
|
... | ... |
@@ -1037,13 +322,13 @@ ClamAV contains HTML normalization code which makes it easier to write signature |
1037 | 1037 |
|
1038 | 1038 |
- javascript - any script contents are normalized and the results appended to this file |
1039 | 1039 |
|
1040 |
-The code automatically decodes JScript.encode parts and char ref’s (e.g. `f`). To create a successful signature for the input file type, the rule must match on the contents of one of the created files. Signatures matching on normalized HTML should have a target type of 3. |
|
1040 |
+The code automatically decodes JScript.encode parts and char ref’s (e.g. `f`). To create a successful signature for the input file type, the rule must match on the contents of one of the created files. Signatures matching on normalized HTML should have a target type of 3. For reference, see [Target Types](Signatures/FileTypes.md#Target-Types). |
|
1041 | 1041 |
|
1042 |
-### Text files |
|
1042 |
+#### Text files |
|
1043 | 1043 |
|
1044 |
-Similarly to HTML all ASCII text files get normalized (converted to lower-case, all superfluous white space and control characters removed, etc.) before scanning. Running `sigtool --ascii-normalise` on a text file will result in a normalized version being written to the file named 'normalised\_text'. Rules matching on normalized ASCII text should have a target type of 7. |
|
1044 |
+Similarly to HTML all ASCII text files get normalized (converted to lower-case, all superfluous white space and control characters removed, etc.) before scanning. Running `sigtool --ascii-normalise` on a text file will result in a normalized version being written to the file named 'normalised\_text'. Rules matching on normalized ASCII text should have a target type of 7. For reference, see [Target Types](Signatures/FileTypes.md#Target-Types). |
|
1045 | 1045 |
|
1046 |
-### Compressed Portable Executable files |
|
1046 |
+#### Compressed Portable Executable files |
|
1047 | 1047 |
|
1048 | 1048 |
If the file is compressed with UPX, FSG, Petite or another PE packer supported by libclamav, ClamAV will attempt to automatically unpack the executable and evaluate signatures against the unpacked executable. To inspect the executable that results from ClamAV's unpacking process, run `clamscan` with `--debug --leave-temps`. Example output for a FSG compressed file: |
1049 | 1049 |
|
... | ... |
@@ -1057,31 +342,32 @@ LibClamAV debug: FSG: Unpacked and rebuilt executable saved in |
1057 | 1057 |
|
1058 | 1058 |
In the example above, `/tmp/clamav-f592b20f9329ac1c91f0e12137bcce6c` is the unpacked executable, and a signature can be written based off of this file. |
1059 | 1059 |
|
1060 |
-## Using sigtool |
|
1061 |
-sigtool pulls in libclamav and provides shortcuts to doing tasks that clamscan does behind the scenes. These can be really useful when writing a signature or trying to get information about a signature that might be causing FPs or performance problems. |
|
1060 |
+### Using `sigtool` |
|
1061 |
+ |
|
1062 |
+`sigtool` pulls in libclamav and provides shortcuts to doing tasks that `clamscan` does behind the scenes. These can be really useful when writing a signature or trying to get information about a signature that might be causing FPs or performance problems. |
|
1062 | 1063 |
|
1063 |
-The following sigtool flags can be especially useful for signature writing: |
|
1064 |
+The following `sigtool` flags can be especially useful for signature writing: |
|
1064 | 1065 |
|
1065 |
-- `--md5` / `--sha1` / `--sha256`: Generate the MD5/SHA1/SHA256 hash and calculate the file size, outputting both as a properly-formatted .hdb/.hsb signature |
|
1066 |
+- `--md5` / `--sha1` / `--sha256`: Generate the MD5/SHA1/SHA256 hash and calculate the file size, outputting both as a properly-formatted `.hdb`/`.hsb` signature |
|
1066 | 1067 |
|
1067 |
-- `--mdb`: Generate section hashes of the specified file. This is useful when generating .mdb signatures. |
|
1068 |
+- `--mdb`: Generate section hashes of the specified file. This is useful when generating `.mdb` signatures. |
|
1068 | 1069 |
|
1069 |
-- `--decode`: Given a ClamAV signature from STDIN, show a more user-friendly representation of it. An example usage of this flag is `cat test.lsb | sigtool --decode`. |
|
1070 |
+- `--decode`: Given a ClamAV signature from STDIN, show a more user-friendly representation of it. An example usage of this flag is `cat test.ldb | sigtool --decode`. |
|
1070 | 1071 |
|
1071 | 1072 |
- `--hex-dump`: Given a sequence of bytes from STDIN, print the hex equivalent. An example usage of this flag is `echo -n "Match on this" | sigtool --hex-dump`. |
1072 | 1073 |
|
1073 |
-- `--html-normalise`: Normalize the specified HTML file in the way that clamscan will before looking for rule matches. Writing signatures off of these files makes it easier to write rules for target type HTML (you'll know what white space, capitalization, etc. to expect). See the [HTML](#html) section for more details. |
|
1074 |
+- `--html-normalise`: Normalize the specified HTML file in the way that `clamscan` will before looking for rule matches. Writing signatures off of these files makes it easier to write rules for target type HTML (you'll know what white space, capitalization, etc. to expect). See the [HTML](#html) section for more details. |
|
1074 | 1075 |
|
1075 |
-- `--ascii-normalise`: Normalize the specified ASCII text file in the way that clamscan will before looking for rule matches. Writing signatures off of this normalized file data makes it easier to write rules for target type Txt (you'll know what white space, capitalization, etc. to expect). See the [Text files](#text-files) sectino for more details. |
|
1076 |
+- `--ascii-normalise`: Normalize the specified ASCII text file in the way that `clamscan` will before looking for rule matches. Writing signatures off of this normalized file data makes it easier to write rules for target type Txt (you'll know what white space, capitalization, etc. to expect). See the [Text files](#text-files) sectino for more details. |
|
1076 | 1077 |
|
1077 | 1078 |
- `--print-certs`: Print the Authenticode signatures of any PE files specified. |
1078 |
- This is useful when writing signature-based .crb rule files. |
|
1079 |
+ This is useful when writing signature-based `.crb` rule files. |
|
1079 | 1080 |
|
1080 | 1081 |
- `--vba`: Extract VBA/Word6 macro code |
1081 | 1082 |
|
1082 | 1083 |
- `--test-sigs`: Given a signature and a sample, determine whether the signature matches and, if so, display the offset into the file where the match occurred. This can be useful for investigating false positive matches in clean files. |
1083 | 1084 |
|
1084 |
-## Inspecting signatures inside a CVD file |
|
1085 |
+### Inspecting signatures inside a CVD file |
|
1085 | 1086 |
|
1086 | 1087 |
CVD (ClamAV Virus Database) is a digitally signed container that includes signature databases in various text formats. The header of the container is a 512 bytes long string with colon separated fields: |
1087 | 1088 |
|
... | ... |
@@ -1106,9 +392,9 @@ eaPVNe2ccXLfEegoda4xU2TezbGfbSEGoU1qolyQYLX674sNA2Ni6l6/CEKYYh |
1106 | 1106 |
Verification OK. |
1107 | 1107 |
``` |
1108 | 1108 |
|
1109 |
-The ClamAV project distributes a number of CVD files, including *main.cvd* and *daily.cvd*. |
|
1109 |
+The ClamAV project distributes a number of CVD files, including `main.cvd` and `daily.cvd`. |
|
1110 | 1110 |
|
1111 |
-To view the signature associated with a given detection name, the CVD files can be unpacked and the underlying text files searched for a rule definition using a tool like `grep`. To do this, use sigtool's `--unpack` flag as follows: |
|
1111 |
+To view the signature associated with a given detection name, the CVD files can be unpacked and the underlying text files searched for a rule definition using a tool like `grep`. To do this, use `sigtool`'s `--unpack` flag as follows: |
|
1112 | 1112 |
|
1113 | 1113 |
```bash |
1114 | 1114 |
$ mkdir /tmp/clamav-sigs |
... | ... |
@@ -1119,8 +405,8 @@ COPYING main.fp main.hsb main.mdb main.ndb |
1119 | 1119 |
main.crb main.hdb main.info main.msb main.sfp |
1120 | 1120 |
``` |
1121 | 1121 |
|
1122 |
-## External tools |
|
1122 |
+### External tools |
|
1123 | 1123 |
|
1124 | 1124 |
Below are tools that can be helpful when writing ClamAV signatures: |
1125 | 1125 |
|
1126 |
- - [CASC](https://github.com/Cisco-Talos/CASC) - CASC is a plugin for IDA Pro that allows the user to highlight sections of code and create a signature based on the underlying instructions (with options to ignore bytes associated with registers, addresses, and offsets). It also contains SigAlyzer, a tool to take an existing signature and locate the regions within the binary that match the subsignatures. |
|
1126 |
+- [CASC](https://github.com/Cisco-Talos/CASC) - CASC is a plugin for IDA Pro that allows the user to highlight sections of code and create a signature based on the underlying instructions (with options to ignore bytes associated with registers, addresses, and offsets). It also contains SigAlyzer, a tool to take an existing signature and locate the regions within the binary that match the subsignatures. |
1127 | 1127 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,34 @@ |
0 |
+# Trusted and Revoked Certificates |
|
1 |
+ |
|
2 |
+Clamav 0.98 checks signed PE files for certificates and verifies each certificate in the chain against a database of trusted and revoked certificates. The signature format is |
|
3 |
+ |
|
4 |
+``` |
|
5 |
+ Name;Trusted;Subject;Serial;Pubkey;Exponent;CodeSign;TimeSign;CertSign; |
|
6 |
+ NotBefore;Comment[;minFL[;maxFL]] |
|
7 |
+``` |
|
8 |
+ |
|
9 |
+where the corresponding fields are: |
|
10 |
+ |
|
11 |
+- `Name:` name of the entry |
|
12 |
+ |
|
13 |
+- `Trusted:` bit field, specifying whether the cert is trusted. 1 for trusted. 0 for revoked |
|
14 |
+ |
|
15 |
+- `Subject:` sha1 of the Subject field in hex |
|
16 |
+ |
|
17 |
+- `Serial:` the serial number as clamscan –debug –verbose reports |
|
18 |
+ |
|
19 |
+- `Pubkey:` the public key in hex |
|
20 |
+ |
|
21 |
+- `Exponent:` the exponent in hex. Currently ignored and hardcoded to 010001 (in hex) |
|
22 |
+ |
|
23 |
+- `CodeSign:` bit field, specifying whether this cert can sign code. 1 for true, 0 for false |
|
24 |
+ |
|
25 |
+- `TimeSign:` bit field. 1 for true, 0 for false |
|
26 |
+ |
|
27 |
+- `CertSign:` bit field, specifying whether this cert can sign other certs. 1 for true, 0 for false |
|
28 |
+ |
|
29 |
+- `NotBefore:` integer, cert should not be added before this variable. Defaults to 0 if left empty |
|
30 |
+ |
|
31 |
+- `Comment:` comments for this entry |
|
32 |
+ |
|
33 |
+The signatures for certs are stored inside `.crb` files. |
0 | 34 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,90 @@ |
0 |
+# Body-based Signature Content Format |
|
1 |
+ |
|
2 |
+ClamAV stores all body-based signatures in a hexadecimal format. In this section by a hex-signature we mean a fragment of malware’s body converted into a hexadecimal string which can be additionally extended using various wildcards. |
|
3 |
+ |
|
4 |
+## Hexadecimal format |
|
5 |
+ |
|
6 |
+You can use `sigtool --hex-dump` to convert any data into a hex-string: |
|
7 |
+ |
|
8 |
+```bash |
|
9 |
+zolw@localhost:/tmp/test$ sigtool --hex-dump |
|
10 |
+How do I look in hex? |
|
11 |
+486f7720646f2049206c6f6f6b20696e206865783f0a |
|
12 |
+``` |
|
13 |
+ |
|
14 |
+## Wildcards |
|
15 |
+ |
|
16 |
+ClamAV supports the following wildcards for hex-signatures: |
|
17 |
+ |
|
18 |
+- `??` |
|
19 |
+ |
|
20 |
+ Match any byte. |
|
21 |
+ |
|
22 |
+- `a?` |
|
23 |
+ |
|
24 |
+ Match a high nibble (the four high bits). |
|
25 |
+ |
|
26 |
+- `?a` |
|
27 |
+ |
|
28 |
+ Match a low nibble (the four low bits). |
|
29 |
+ |
|
30 |
+- `*` |
|
31 |
+ |
|
32 |
+ Match any number of bytes. |
|
33 |
+ |
|
34 |
+- `{n}` |
|
35 |
+ |
|
36 |
+ Match `n` bytes. |
|
37 |
+ |
|
38 |
+- `{-n}` |
|
39 |
+ |
|
40 |
+ Match `n` or less bytes. |
|
41 |
+ |
|
42 |
+- `{n-}` |
|
43 |
+ |
|
44 |
+ Match `n` or more bytes. |
|
45 |
+ |
|
46 |
+- `{n-m}` |
|
47 |
+ |
|
48 |
+ Match between `n` and `m` bytes (where `m > n`). |
|
49 |
+ |
|
50 |
+- `HEXSIG[x-y]aa` or `aa[x-y]HEXSIG` |
|
51 |
+ |
|
52 |
+ Match `aa` anchored to a hex-signature, see [Bugzilla ticket 776](https://bugzilla.clamav.net/show_bug.cgi?id=776) for discussion and |
|
53 |
+ examples. |
|
54 |
+ |
|
55 |
+The range signatures `*` and `{}` virtually separate a hex-signature into two parts, eg. `aabbcc*bbaacc` is treated as two sub-signatures `aabbcc` and `bbaacc` with any number of bytes between them. It’s a requirement that each sub-signature includes a block of two static characters somewhere in its body. Note that there is one exception to this restriction; that is when the range wildcard is of the form `{n}` with `n<128`. In this case, ClamAV uses an optimization and translates `{n}` to the string consisting of `n ??` character wildcards. Character wildcards do not divide hex signatures into two parts and so the two static character requirement does not apply. |
|
56 |
+ |
|
57 |
+## Character classes |
|
58 |
+ |
|
59 |
+ClamAV supports the following character classes for hex-signatures: |
|
60 |
+ |
|
61 |
+- `(B)` |
|
62 |
+ |
|
63 |
+ Match word boundary (including file boundaries). |
|
64 |
+ |
|
65 |
+- `(L)` |
|
66 |
+ |
|
67 |
+ Match CR, CRLF or file boundaries. |
|
68 |
+ |
|
69 |
+- `(W)` |
|
70 |
+ |
|
71 |
+ Match a non-alphanumeric character. |
|
72 |
+ |
|
73 |
+## Alternate strings |
|
74 |
+ |
|
75 |
+- Single-byte alternates (clamav-0.96) `(aa|bb|cc|...)` or `!(aa|bb|cc|...)` Match a member from a set of bytes (eg: `aa`, `bb`, `cc`, ...). |
|
76 |
+ - Negation operation can be applied to match any non-member, assumed to be one-byte in length. |
|
77 |
+ - Signature modifiers and wildcards cannot be applied. |
|
78 |
+ |
|
79 |
+- Multi-byte fixed length alternates `(aaaa|bbbb|cccc|...)` or `!(aaaa|bbbb|cccc|...)` Match a member from a set of multi-byte alternates (eg: aaaa, bbbb, cccc, ...) of n-length. |
|
80 |
+ - All set members must be the same length. |
|
81 |
+ - Negation operation can be applied to match any non-member, assumed to be n-bytes in length (clamav-0.98.2). |
|
82 |
+ - Signature modifiers and wildcards cannot be applied. |
|
83 |
+ |
|
84 |
+- Generic alternates (clamav-0.99) `(alt1|alt2|alt3|...)` Match a member from a set of alternates (eg: alt1, alt2, alt3, ...) that can be of variable lengths. |
|
85 |
+ - Negation operation cannot be applied. |
|
86 |
+ - Signature modifiers and nibble wildcards (eg: `??, a?, ?a`) can be applied. |
|
87 |
+ - Ranged wildcards (eg: `{n-m}`) are limited to a fixed range of less than 128 bytes (eg: `{1} -> {127}`). |
|
88 |
+ |
|
89 |
+Note that using signature modifiers and wildcards classifies the alternate type to be a generic alternate. Thus single-byte alternates and multi-byte fixed length alternates can use signature modifiers and wildcards but will be classified as generic alternate. This means that negation cannot be applied in this situation and there is a slight performance impact. |
0 | 90 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,11 @@ |
0 |
+# Bytecode Signatures |
|
1 |
+ |
|
2 |
+Bytecode Signatures are the means by which more complex matching can be performed by writing C code to parse sample content at various stages in file extraction. |
|
3 |
+ |
|
4 |
+It is less complicated than it sounds. Essentially the signature author writes a function in C is compiled down to an intermediate language called "bytecode". This bytecode is encoded in ASCII `.cbc` file and distributed in `bytecode.[cvd|cld]`. When the database is loaded, ClamAV can interpret this bytecode to execute the function. |
|
5 |
+ |
|
6 |
+Bytecode functions are provided with a set of API's that may be used to access the sample data, and to access what metadata ClamAV already has concerning the sample. |
|
7 |
+ |
|
8 |
+The function may at any time call an API to flag the sample as malicious, and may provide the signature/virus name at that time. This means a single bytecode signature (function) is written to handle a given file type and may trigger different alerts with different signature names as additional malicious characteristics for the file type are identified. That isn't to say that only one bytecode signature may be assigned to a given filetype, but that a single author may find it to be more efficient to use a bytecode signature to identify more than one type of malware. |
|
9 |
+ |
|
10 |
+The specifics on how to write and compile bytecode signatures are outside of the scope of this documentation. Extensive documentation on ClamAV Bytecode Signatures are provided with the [ClamAV Bytecode Compiler](https://github.com/vrtadmin/clamav-bytecode-compiler). |
0 | 11 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,44 @@ |
0 |
+# Signatures based on container metadata |
|
1 |
+ |
|
2 |
+ClamAV 0.96 allows creating generic signatures matching files stored inside different container types which meet specific conditions. The signature format is: |
|
3 |
+ |
|
4 |
+``` |
|
5 |
+ VirusName:ContainerType:ContainerSize:FileNameREGEX: |
|
6 |
+ FileSizeInContainer:FileSizeReal:IsEncrypted:FilePos: |
|
7 |
+ Res1:Res2[:MinFL[:MaxFL]] |
|
8 |
+``` |
|
9 |
+ |
|
10 |
+where the corresponding fields are: |
|
11 |
+ |
|
12 |
+- `VirusName:` Virus name to be displayed when signature matches. |
|
13 |
+ |
|
14 |
+- `ContainerType:` The file type containing the target file. For example: |
|
15 |
+ - `CL_TYPE_ZIP`, |
|
16 |
+ - `CL_TYPE_RAR`, |
|
17 |
+ - `CL_TYPE_ARJ`, |
|
18 |
+ - `CL_TYPE_MSCAB`, |
|
19 |
+ - `CL_TYPE_7Z`, |
|
20 |
+ - `CL_TYPE_MAIL`, |
|
21 |
+ - `CL_TYPE_(POSIX|OLD)_TAR`, |
|
22 |
+ - `CL_TYPE_CPIO_(OLD|ODC|NEWC|CRC)` |
|
23 |
+ |
|
24 |
+ Use `*` as a wild card to indicate that container type may be any file type. |
|
25 |
+ For a full list of ClamAV file types, see the [ClamAV File Types Reference](ClamAVFileTypes.md). |
|
26 |
+ |
|
27 |
+- `ContainerSize:` size of the container file itself (eg. size of the zip archive) specified in bytes as absolute value or range `x-y`. |
|
28 |
+ |
|
29 |
+- `FileNameREGEX:` regular expression describing name of the target file |
|
30 |
+ |
|
31 |
+- `FileSizeInContainer:` usually compressed size; for MAIL, TAR and CPIO == `FileSizeReal`; specified in bytes as absolute value or range. |
|
32 |
+ |
|
33 |
+- `FileSizeReal:` usually uncompressed size; for MAIL, TAR and CPIO == `FileSizeInContainer`; absolute value or range. |
|
34 |
+ |
|
35 |
+- `IsEncrypted:` 1 if the target file is encrypted, 0 if it’s not and `*` to ignore |
|
36 |
+ |
|
37 |
+- `FilePos:` file position in container (counting from 1); absolute value or range. |
|
38 |
+ |
|
39 |
+- `Res1:` when `ContainerType` is `CL_TYPE_ZIP` or `CL_TYPE_RAR` this field is treated as a CRC sum of the target file specified in hexadecimal format; for other container types it’s ignored. |
|
40 |
+ |
|
41 |
+- `Res2:` not used as of ClamAV 0.96. |
|
42 |
+ |
|
43 |
+The signatures for container files are stored inside `.cdb` files. |
0 | 44 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,15 @@ |
0 |
+# Database Info |
|
1 |
+ |
|
2 |
+The `.info` file format specifies information about the other database files unpacked from a CVD or CLD database archive. This file exists for the purposes of validating the correctness of the official ClamAV database container files and cannot be loaded a la carte. |
|
3 |
+ |
|
4 |
+The format is simply: |
|
5 |
+ |
|
6 |
+``` |
|
7 |
+name:size:sha256 |
|
8 |
+``` |
|
9 |
+ |
|
10 |
+`name`: The database file name. |
|
11 |
+ |
|
12 |
+`size`: The size in bytes of the database. |
|
13 |
+ |
|
14 |
+`sha256`: A SHA256 hash of the database. |
0 | 15 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,81 @@ |
0 |
+# Dynamic Configuration (DCONF) |
|
1 |
+ |
|
2 |
+ClamAV supports a limited set of configuration options that may be enabled or disabled via settings in the `*.cfg` database. At this time, these settings are distributed in `daily.cfg`. |
|
3 |
+ |
|
4 |
+The goal of DCONF is to enable the ClamAV team to rapidly disable new or experimental features for specific ClamAV versions if a significant defect is discovered after release. |
|
5 |
+ |
|
6 |
+This database is small, and the settings are largely vestigial. The team has not had a need to disable many features in a long time, and so the ClamAV versions in the settings at this time should no longer be in use. |
|
7 |
+ |
|
8 |
+The strings and values referenced in `daily.cfg` are best cross-referenced with the macros and structures defined here: |
|
9 |
+ |
|
10 |
+* https://github.com/Cisco-Talos/clamav-devel/blob/dev/0.101/libclamav/dconf.h#L49 |
|
11 |
+* https://github.com/Cisco-Talos/clamav-devel/blob/dev/0.101/libclamav/dconf.c#L54 |
|
12 |
+ |
|
13 |
+The format for a DCONF signature is: |
|
14 |
+ |
|
15 |
+``` |
|
16 |
+Category:Flags:StartFlevel:EndFlevel |
|
17 |
+``` |
|
18 |
+ |
|
19 |
+`Category` may be one of: |
|
20 |
+ |
|
21 |
+* PE |
|
22 |
+* ELF |
|
23 |
+* MACHO |
|
24 |
+* ARCHIVE |
|
25 |
+* DOCUMENT |
|
26 |
|
|
27 |
+* OTHER |
|
28 |
+* PHISHING |
|
29 |
+* BYTECODE |
|
30 |
+* STATS |
|
31 |
+* PCRE |
|
32 |
+ |
|
33 |
+`Flags`: |
|
34 |
+ |
|
35 |
+Every feature that may be configured via DCONF is listed in `struct dconf_module modules` in `libclamav/dconf.c`. Any given feature may be default-on or default-off. Default-on features have the 4th field set to a `1` and default off are set to `0`. The `Flags` field for a given `Category` overrides the defaults for all of the options listed under that category. |
|
36 |
+ |
|
37 |
+A settings of `0x0`, for example, means that all options the category be disabled. |
|
38 |
+ |
|
39 |
+The macros listed in `libclamav/dconf.h` will help you identify which bits to set to get the desired results. |
|
40 |
+ |
|
41 |
+`StartFlevel`: |
|
42 |
+ |
|
43 |
+This is the [FLEVEL](FunctionalityLevels.md) of the minimum ClamAV engine for which you want the settings to be in effect. |
|
44 |
+ |
|
45 |
+`EndFlevel`: |
|
46 |
+ |
|
47 |
+This is the [FLEVEL](FunctionalityLevels.md) of the maximum ClamAV engine for which you want the settings to be in effect. You may wish to select `255` to override the defaults of future releases. |
|
48 |
+ |
|
49 |
+## Example |
|
50 |
+ |
|
51 |
+Consider the `OTHER_CONF_PDFNAMEOBJ` option in the `category` `OTHER`. |
|
52 |
+ |
|
53 |
+```c |
|
54 |
+#define OTHER_CONF_UUENC 0x1 // Default: 1 |
|
55 |
+#define OTHER_CONF_SCRENC 0x2 // Default: 1 |
|
56 |
+#define OTHER_CONF_RIFF 0x4 // Default: 1 |
|
57 |
+#define OTHER_CONF_JPEG 0x8 // Default: 1 |
|
58 |
+#define OTHER_CONF_CRYPTFF 0x10 // Default: 1 |
|
59 |
+#define OTHER_CONF_DLP 0x20 // Default: 1 |
|
60 |
+#define OTHER_CONF_MYDOOMLOG 0x40 // Default: 1 |
|
61 |
+#define OTHER_CONF_PREFILTERING 0x80 // Default: 1 |
|
62 |
+#define OTHER_CONF_PDFNAMEOBJ 0x100 // Default: 1 |
|
63 |
+#define OTHER_CONF_PRTNINTXN 0x200 // Default: 1 |
|
64 |
+#define OTHER_CONF_LZW 0x400 // Default: 1 |
|
65 |
+``` |
|
66 |
+ |
|
67 |
+All of the `OTHER` options, including `OTHER_CONF_PDFNAMEOBJ` are default-on. To disable the option for ClamAV v0.100.X but leave the other options in their default settings, we would need to set the flags to: |
|
68 |
+ |
|
69 |
+```binary |
|
70 |
+0110 1111 1111 |
|
71 |
+ ^pdfnameobj off |
|
72 |
+``` |
|
73 |
+ |
|
74 |
+Or in hex: `0x6FF` |
|
75 |
+ |
|
76 |
+The example setting to place in `daily.cfg` then woudl be: |
|
77 |
+ |
|
78 |
+``` |
|
79 |
+OTHER:0x6FF:90:99 |
|
80 |
+``` |
0 | 81 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,23 @@ |
0 |
+# Passwords for archive files \[experimental\] |
|
1 |
+ |
|
2 |
+ClamAV 0.99 allows for users to specify password attempts for certain password-compatible archives. Passwords will be attempted in order of appearance in the password signature file which use the extension of `.pwdb`. If no passwords apply or none are provided, ClamAV will default to the original behavior of parsing the file. Currently, as of ClamAV 0.99 \[flevel 81\], only `.zip` archives using the traditional PKWARE encryption are supported. The signature format is |
|
3 |
+ |
|
4 |
+``` |
|
5 |
+SignatureName;TargetDescriptionBlock;PWStorageType;Password |
|
6 |
+``` |
|
7 |
+ |
|
8 |
+where: |
|
9 |
+ |
|
10 |
+- `SignatureName`: name to be displayed during debug when a password is successful |
|
11 |
+ |
|
12 |
+- `TargetDescriptionBlock`: provides information about the engine and target file with comma separated Arg:Val pairs |
|
13 |
+ - `Engine:X-Y`: Required engine functionality level. See the [FLEVEL reference](FunctionalityLevels.md) for details. |
|
14 |
+ - `Container:CL_TYPE_*`: File type of applicable containers |
|
15 |
+ |
|
16 |
+- `PWStorageType`: determines how the password field is parsed |
|
17 |
+ - 0 = cleartext |
|
18 |
+ - 1 = hex |
|
19 |
+ |
|
20 |
+- `Password`: value used in password attempt |
|
21 |
+ |
|
22 |
+The signatures for password attempts are stored inside `.pwdb` files. |
0 | 23 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,37 @@ |
0 |
+# Extended signature format |
|
1 |
+ |
|
2 |
+The extended signature format is ClamAV's most basic type of body-based signature since the deprecation of the original `.db` database format. |
|
3 |
+ |
|
4 |
+Extended sigantures allow for specification of additional information beyond just hexidecimal content such as a file "target type", virus offset, or engine functionality level (FLEVEL), making the detection more reliable. |
|
5 |
+ |
|
6 |
+The format is: |
|
7 |
+ |
|
8 |
+``` |
|
9 |
+ MalwareName:TargetType:Offset:HexSignature[:min_flevel:[max_flevel]] |
|
10 |
+``` |
|
11 |
+ |
|
12 |
+`MalwareName`: The virus name. Should conform to the standards defined [here](../Signatures.md#Signature-names). |
|
13 |
+ |
|
14 |
+`TargetType`: A number specifying the type of the target file: [Target Types](FileTypes.md#Target-Types) |
|
15 |
+ |
|
16 |
+`Offset`: An asterisk or a decimal number `n` possibly combined with a special modifier: |
|
17 |
+ |
|
18 |
+- `*` = any |
|
19 |
+- `n` = absolute offset |
|
20 |
+- `EOF-n` = end of file minus `n` bytes |
|
21 |
+ |
|
22 |
+Signatures for PE, ELF and Mach-O files additionally support: |
|
23 |
+ |
|
24 |
+- `EP+n` = entry point plus n bytes (`EP+0` for `EP`) |
|
25 |
+- `EP-n` = entry point minus n bytes |
|
26 |
+- `Sx+n` = start of section `x`’s (counted from 0) data plus `n` bytes |
|
27 |
+- `SEx` = entire section `x` (offset must lie within section boundaries) |
|
28 |
+- `SL+n` = start of last section plus `n` bytes |
|
29 |
+ |
|
30 |
+All the above offsets except `*` can be turned into **floating offsets** and represented as `Offset,MaxShift` where `MaxShift` is an unsigned integer. A floating offset will match every offset between `Offset` and `Offset+MaxShift`, eg. `10,5` will match all offsets from 10 to 15 and `EP+n,y` will match all offsets from `EP+n` to `EP+n+y`. Versions of ClamAV older than 0.91 will silently ignore the `MaxShift` extension and only use `Offset`. Optional `MinFL` and `MaxFL` parameters can restrict the signature to specific engine releases. All signatures in the extended format must be placed inside `*.ndb` files. |
|
31 |
+ |
|
32 |
+`HexSignature`: The body-based content matching [format](BodySignatureFormat.md). |
|
33 |
+ |
|
34 |
+`min_flevel`: (optional) The minimum ClamAV engine that the file type signature works with. See the [FLEVEL reference](FunctionalityLevels.md) for details. To be used in the event that file type support has been recently added. |
|
35 |
+ |
|
36 |
+`max_flevel`: (optional, requires `min_flevel`) The maximum ClamAV engine that the file type signature works with. To be used in the event that file type support has been recently removed. |
0 | 37 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,33 @@ |
0 |
+# File Type Magic |
|
1 |
+ |
|
2 |
+ClamAV's primary mechanism for determining file types is to match the file with a File Type Magic signature. These file type signatures are compiled into ClamAV, and may also be overridden dynamically using the definition founds found in a `*.ftm` file. |
|
3 |
+ |
|
4 |
+The ClamAV standard signature database includes these definitions in `daily.ftm`. |
|
5 |
+ |
|
6 |
+The signature format is not too disimilar from NDB body-based signatures. |
|
7 |
+ |
|
8 |
+The format is: |
|
9 |
+ |
|
10 |
+``` |
|
11 |
+ magictype:offset:magicbytes:name:type:type[:min_flevel[:max_flevel]] |
|
12 |
+``` |
|
13 |
+ |
|
14 |
+Where: |
|
15 |
+ |
|
16 |
+`magictype`: Supported magic types include: |
|
17 |
+ |
|
18 |
+* 0 - direct memory comparison of `magicbytes` for file types |
|
19 |
+* 1 - The `magicbytes` use the body-based content matching [format](BodySignatureFormat.md). |
|
20 |
+* 4 - direct memory comparison of `magicbytes` for partition types (HFS+, HFSX) |
|
21 |
+ |
|
22 |
+`offset`: The offset from start of the file to match against. May be `*` if `magictype` is 1. |
|
23 |
+ |
|
24 |
+`name`: A descriptive name for the file type. |
|
25 |
+ |
|
26 |
+`rtype`: Usually CL_TYPE_ANY. |
|
27 |
+ |
|
28 |
+`type`: The CL_TYPE corresponding with the file type signature. See the [CL_TYPE reference](ClamAVFileTypes.md) for details. |
|
29 |
+ |
|
30 |
+`min_flevel`: (optional) The minimum ClamAV engine that the file type signature works with. See the [FLEVEL reference](FunctionalityLevels.md) for details. To be used in the event that file type support has been recently added. |
|
31 |
+ |
|
32 |
+`max_flevel`: (optional, requires `min_flevel`) The maximum ClamAV engine that the file type signature works with. To be used in the event that file type support has been recently removed. |
0 | 33 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,120 @@ |
0 |
+# ClamAV File Types |
|
1 |
+ |
|
2 |
+ClamAV maintains it's own file typing format and assigns these types using either: |
|
3 |
+ |
|
4 |
+- Evaluation of a unique sequence of bytes at the start of a file ([File Type Magic](Signatures/FileTypeMagic.md)). |
|
5 |
+- File type indicators when parsing container files. |
|
6 |
+ - For example: |
|
7 |
+ CL_TYPE_SCRIPT may be assigned to data contained in a PDF when the PDF indicates that a stream of bytes is "Javascript" |
|
8 |
+- File type determination based on the names or characteristics contained within the file. |
|
9 |
+ - For example: |
|
10 |
+ CL_TYPE_OOXML_WORD may be assigned to a Zip file containing files with specific names. |
|
11 |
+ |
|
12 |
+## Target Types |
|
13 |
+ |
|
14 |
+A Target Type is an integer that indicates which kind of file the signature will match against. Target Type notation was first created for the purposes writing efficient signatures. A signature with a target type of `0` will be run against every file type, and thus is not ideal. However, the Target Type notation is limited and it may be unavoidable. |
|
15 |
+ |
|
16 |
+Although the newer CL_TYPE string name notation has replaced the Target Type for some signature formats, many signature formats require a target type number. |
|
17 |
+ |
|
18 |
+This is the current list of available Targe Types: |
|
19 |
+ |
|
20 |
+- 0 = any file |
|
21 |
+- 1 = Portable Executable, both 32- and 64-bit. |
|
22 |
+- 2 = OLE2 containers, including their specific macros. The OLE2 format is primarily used by MS Office and MSI installation files. |
|
23 |
+- 3 = HTML (normalized) |
|
24 |
+- 4 = Mail file |
|
25 |
+- 5 = Graphics |
|
26 |
+- 6 = ELF |
|
27 |
+- 7 = ASCII text file (normalized) |
|
28 |
+- 8 = Unused |
|
29 |
+- 9 = Mach-O files |
|
30 |
+- 10 = PDF files |
|
31 |
+- 11 = Flash files |
|
32 |
+- 12 = Java class files |
|
33 |
+ |
|
34 |
+**_Important_: HTML, ASCII, Javascript are all normalized. |
|
35 |
+ |
|
36 |
+- ASCII: |
|
37 |
+ - All lowercase. |
|
38 |
+- HTML: |
|
39 |
+ - Whitespace transformed to spaces, tags/tag attributes normalized, all lowercase. |
|
40 |
+- Javascript: |
|
41 |
+ - All strings are normalized (hex encoding is decoded), numbers are parsed and normalized, local variables/function names are normalized to ’n001’ format, argument to eval() is parsed as JS again, unescape() is handled, some simple JS packers are handled, output is whitespace normalized. |
|
42 |
+ |
|
43 |
+## CL_TYPEs |
|
44 |
+ |
|
45 |
+ClamAV Types are prefixed with `CL_TYPE_`. The following is an exhaustive list of all current CL_TYPE's. |
|
46 |
+ |
|
47 |
+| CL_TYPE | Description | |
|
48 |
+|------------------------|--------------------------------------------------------------| |
|
49 |
+| `CL_TYPE_7Z` | 7-Zip Archive | |
|
50 |
+| `CL_TYPE_7ZSFX` | Self-Extracting 7-Zip Archive | |
|
51 |
+| `CL_TYPE_APM` | Disk Image - Apple Partition Map | |
|
52 |
+| `CL_TYPE_ARJ` | ARJ Archive | |
|
53 |
+| `CL_TYPE_ARJSFX` | Self-Extracting ARJ Archive | |
|
54 |
+| `CL_TYPE_AUTOIT` | AutoIt Automation Executable | |
|
55 |
+| `CL_TYPE_BINARY_DATA` | binary data | |
|
56 |
+| `CL_TYPE_BINHEX` | BinHex Macintosh 7-bit ASCII email attachment encoding | |
|
57 |
+| `CL_TYPE_BZ` | BZip Compressed File | |
|
58 |
+| `CL_TYPE_CABSFX` | Self-Extracting Microsoft CAB Archive | |
|
59 |
+| `CL_TYPE_CPIO_CRC` | CPIO Archive (CRC) | |
|
60 |
+| `CL_TYPE_CPIO_NEWC` | CPIO Archive (NEWC) | |
|
61 |
+| `CL_TYPE_CPIO_ODC` | CPIO Archive (ODC) | |
|
62 |
+| `CL_TYPE_CPIO_OLD` | CPIO Archive (OLD, Little Endian or Big Endian) | |
|
63 |
+| `CL_TYPE_CRYPTFF` | Files encrypted by CryptFF malware | |
|
64 |
+| `CL_TYPE_DMG` | Apple DMG Archive | |
|
65 |
+| `CL_TYPE_ELF` | ELF Executable (Linux/Unix program or library) | |
|
66 |
+| `CL_TYPE_GPT` | Disk Image - GUID Partition Table | |
|
67 |
+| `CL_TYPE_GRAPHICS` | TIFF (Little Endian or Big Endian) | |
|
68 |
+| `CL_TYPE_GZ` | GZip Compressed File | |
|
69 |
+| `CL_TYPE_HTML_UTF16` | Wide-Character / UTF16 encoded HTML | |
|
70 |
+| `CL_TYPE_HTML` | HTML data | |
|
71 |
+| `CL_TYPE_HWP3` | Hangul Word Processor (3.X) | |
|
72 |
+| `CL_TYPE_HWPOLE2` | Hangul Word Processor embedded OLE2 | |
|
73 |
+| `CL_TYPE_INTERNAL` | Internal properties | |
|
74 |
+| `CL_TYPE_ISHIELD_MSI` | Windows Install Shield MSI installer | |
|
75 |
+| `CL_TYPE_ISO9660` | ISO 9660 file system for optical disc media | |
|
76 |
+| `CL_TYPE_JAVA` | Java Class File | |
|
77 |
+| `CL_TYPE_LNK` | Microsoft Windows Shortcut File | |
|
78 |
+| `CL_TYPE_MACHO_UNIBIN` | Universal Binary/Java Bytecode | |
|
79 |
+| `CL_TYPE_MACHO` | Apple/NeXTSTEP Mach-O Executable file format | |
|
80 |
+| `CL_TYPE_MAIL` | Email file | |
|
81 |
+| `CL_TYPE_MBR` | Disk Image - Master Boot Record | |
|
82 |
+| `CL_TYPE_MHTML` | MHTML Saved Web Page | |
|
83 |
+| `CL_TYPE_MSCAB` | Microsoft CAB Archive | |
|
84 |
+| `CL_TYPE_MSCHM` | Microsoft CHM help archive | |
|
85 |
+| `CL_TYPE_MSEXE` | Microsoft EXE / DLL Executable file | |
|
86 |
+| `CL_TYPE_MSOLE2` | Microsoft OLE2 Container file | |
|
87 |
+| `CL_TYPE_MSSZDD` | Microsoft Compressed EXE | |
|
88 |
+| `CL_TYPE_NULSFT` | NullSoft Scripted Installer program | |
|
89 |
+| `CL_TYPE_OLD_TAR` | TAR archive (old) | |
|
90 |
+| `CL_TYPE_OOXML_HWP` | Hangul Office Open Word Processor (5.X) | |
|
91 |
+| `CL_TYPE_OOXML_PPT` | Microsoft Office Open XML PowerPoint | |
|
92 |
+| `CL_TYPE_OOXML_WORD` | Microsoft Office Open Word 2007+ | |
|
93 |
+| `CL_TYPE_OOXML_XL` | Microsoft Office Open Excel 2007+ | |
|
94 |
+| `CL_TYPE_PART_HFSPLUS` | Apple HFS+ partition | |
|
95 |
+| `CL_TYPE_PDF` | Adobe PDF document | |
|
96 |
+| `CL_TYPE_POSIX_TAR` | TAR archive | |
|
97 |
+| `CL_TYPE_PS` | Postscript | |
|
98 |
+| `CL_TYPE_RAR` | RAR Archive | |
|
99 |
+| `CL_TYPE_RARSFX` | Self-Extracting RAR Archive | |
|
100 |
+| `CL_TYPE_RIFF` | Resource Interchange File Format container formatted file | |
|
101 |
+| `CL_TYPE_RTF` | Rich Text Format document | |
|
102 |
+| `CL_TYPE_SCRENC` | Files encrypted by ScrEnc malware | |
|
103 |
+| `CL_TYPE_SCRIPT` | Generic type for scripts (Javascript, Python, etc) | |
|
104 |
+| `CL_TYPE_SIS` | Symbian OS Software Installation Script Archive | |
|
105 |
+| `CL_TYPE_SWF` | Adobe Flash File (LZMA, Zlib, or uncompressed) | |
|
106 |
+| `CL_TYPE_TEXT_ASCII` | ASCII text | |
|
107 |
+| `CL_TYPE_TEXT_UTF16BE` | UTF-16BE text | |
|
108 |
+| `CL_TYPE_TEXT_UTF16LE` | UTF-16LE text | |
|
109 |
+| `CL_TYPE_TEXT_UTF8` | UTF-8 text | |
|
110 |
+| `CL_TYPE_TNEF` | Microsoft Outlook & Exchange email attachment format | |
|
111 |
+| `CL_TYPE_UUENCODED` | UUEncoded (Unix-to-Unix) binary file (Unix email attachment) | |
|
112 |
+| `CL_TYPE_XAR` | XAR Archive | |
|
113 |
+| `CL_TYPE_XDP` | Adobe XDP - Embedded PDF | |
|
114 |
+| `CL_TYPE_XML_HWP` | Hangul Word Processor XML (HWPML) Document | |
|
115 |
+| `CL_TYPE_XML_WORD` | Microsoft Word 2003 XML Document | |
|
116 |
+| `CL_TYPE_XML_XL` | Microsoft Excel 2003 XML Document | |
|
117 |
+| `CL_TYPE_XZ` | XZ Archive | |
|
118 |
+| `CL_TYPE_ZIP` | Zip Archive | |
|
119 |
+| `CL_TYPE_ZIPSFX` | Self-Extracting Zip Archive | |
0 | 120 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,30 @@ |
0 |
+# Functionality Levels (FLEVELs) |
|
1 |
+ |
|
2 |
+The Functionality Level (or FLEVEL) is an integer that signatures may use to define which versions of ClamAV the signature features support. It is up to the signature writers to select the correct FLEVEL or range of FLEVELs when writing a signature so that it does not cause failures in older versions of ClamAV. |
|
3 |
+ |
|
4 |
+Setting appropriate FLEVELs in signatures is particularly crucial when using features added in the last 3-4 major release versions. |
|
5 |
+ |
|
6 |
+## ClamAV Version to FLEVEL chart |
|
7 |
+ |
|
8 |
+| flevel | version | release | new signature features | |
|
9 |
+|--------|---------|---------|------------------------------------------------------------------------| |
|
10 |
+| 41 | 0.95.0 | 3/2009 | Ignores use ign format (including line number). | |
|
11 |
+| 51 | 0.96.0 | 3/2010 | Bytecode & CDB sigs. Start using ign2. | |
|
12 |
+| 56 | 0.96.4 | 10/2010 | Min level for bytecode sigs. | |
|
13 |
+| 60 | 0.97.0 | 2/2011 | | |
|
14 |
+| 74 | 0.98.0 | 9/2013 | ISO9660 scanning support. All-match feature. | |
|
15 |
+| | | | Wild card bracket notation{} for body-based signatures. | |
|
16 |
+| | | | "SE" offset modifier. | |
|
17 |
+| | | | Target types 10 - 13: (PDF, (SWF) Flash, Java, Internal). | |
|
18 |
+| 76 | 0.98.1 | 1/2014 | XZ support and ForceToDisk scan option. | |
|
19 |
+| | | | Libxml2, XAR, DMG, HFS+/HFSX. | |
|
20 |
+| | | | FTM type 4 (in-buffer partition magic, analogous to type 0 for files). | |
|
21 |
+| 79 | 0.98.5 | 11/2014 | File properties (preclass). Target type 13: for preclass feature. | |
|
22 |
+| 81 | 0.99.0 | 11/2015 | Yara and PCRE support. Target type 14: non-listed types ("other"). | |
|
23 |
+| 82 | 0.99.1 | 2/2016 | Hangul Word Processor (HWP) type file parser. | |
|
24 |
+| 90 | 0.100 | 4/2018 | "Intermediates" logical sig expression option. | |
|
25 |
+| | | | MHTML and PostScript types. | |
|
26 |
+| | | | Substring wildcard (*) fix: order matters, substrings can't overlap. | |
|
27 |
+| 100 | 0.101 | 12/2018 | "Byte-Compare" Logical subsignature. Windows Shortcut (LNK) type. | |
|
28 |
+ |
|
29 |
+For more inforamtion on ClamAV file type support, see the [File Types Reference](FileTypes.md). |
0 | 30 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,67 @@ |
0 |
+# File hash signatures |
|
1 |
+ |
|
2 |
+The easiest way to create signatures for ClamAV is to use filehash checksums, however this method can be only used against static malware. |
|
3 |
+ |
|
4 |
+## MD5 hash-based signatures |
|
5 |
+ |
|
6 |
+To create a MD5 signature for `test.exe` use the `--md5` option of |
|
7 |
+sigtool: |
|
8 |
+ |
|
9 |
+```bash |
|
10 |
+zolw@localhost:/tmp/test$ sigtool --md5 test.exe > test.hdb |
|
11 |
+zolw@localhost:/tmp/test$ cat test.hdb |
|
12 |
+48c4533230e1ae1c118c741c0db19dfb:17387:test.exe |
|
13 |
+``` |
|
14 |
+ |
|
15 |
+That’s it! The signature is ready for use: |
|
16 |
+ |
|
17 |
+```bash |
|
18 |
+zolw@localhost:/tmp/test$ clamscan -d test.hdb test.exe |
|
19 |
+test.exe: test.exe FOUND |
|
20 |
+ |
|
21 |
+----------- SCAN SUMMARY ----------- |
|
22 |
+Known viruses: 1 |
|
23 |
+Scanned directories: 0 |
|
24 |
+Engine version: 0.92.1 |
|
25 |
+Scanned files: 1 |
|
26 |
+Infected files: 1 |
|
27 |
+Data scanned: 0.02 MB |
|
28 |
+Time: 0.024 sec (0 m 0 s) |
|
29 |
+``` |
|
30 |
+ |
|
31 |
+You can change the name (by default sigtool uses the name of the file) and place it inside a `*.hdb` file. A single database file can include any number of signatures. To get them automatically loaded each time `clamscan`/`clamd` starts just copy the database file(s) into the local virus database directory (eg. `/usr/local/share/clamav`). |
|
32 |
+ |
|
33 |
+*The hash-based signatures shall not be used for text files, HTML and any other data that gets internally preprocessed before pattern matching. If you really want to use a hash signature in such a case, run `clamscan` with `--debug` and `--leave-temps` flags as described above and create a signature for a preprocessed file left in `/tmp`. Please keep in mind that a hash signature will stop matching as soon as a single byte changes in the target file.* |
|
34 |
+ |
|
35 |
+## SHA1 and SHA256 hash-based signatures |
|
36 |
+ |
|
37 |
+ClamAV 0.98 has also added support for SHA1 and SHA256 file checksums. The format is the same as for MD5 file checksum. It can differentiate between them based on the length of the hash string in the signature. For best backwards compatibility, these should be placed inside a `*.hsb` file. The format is: |
|
38 |
+ |
|
39 |
+``` |
|
40 |
+HashString:FileSize:MalwareName |
|
41 |
+``` |
|
42 |
+ |
|
43 |
+## Hash signatures with unknown size |
|
44 |
+ |
|
45 |
+ClamAV 0.98 has also added support for hash signatures where the size is not known but the hash is. It is much more performance-efficient to use signatures with specific sizes, so be cautious when using this feature. For these cases, the ’\*’ character can be used in the size field. To ensure proper backwards compatibility with older versions of ClamAV, these signatures must have a minimum functional level of 73 or higher. Signatures that use the wildcard size without this level set will be rejected as malformed. |
|
46 |
+ |
|
47 |
+Sample .hsb signature matching any size: |
|
48 |
+``` |
|
49 |
+ HashString:*:MalwareName:73 |
|
50 |
+``` |
|
51 |
+Sample .msb signature matching any size: |
|
52 |
+``` |
|
53 |
+ *:PESectionHash:MalwareName:73 |
|
54 |
+``` |
|
55 |
+ |
|
56 |
+## PE section based hash signatures |
|
57 |
+ |
|
58 |
+You can create a hash signature for a specific section in a PE file. Such signatures shall be stored inside `.mdb` (MD5) and `.msb` files in the following format: |
|
59 |
+ |
|
60 |
+``` |
|
61 |
+ PESectionSize:PESectionHash:MalwareName |
|
62 |
+``` |
|
63 |
+ |
|
64 |
+The easiest way to generate MD5 based section signatures is to extract target PE sections into separate files and then run sigtool with the option `--mdb` |
|
65 |
+ |
|
66 |
+ClamAV 0.98 has also added support for SHA1 and SHA256 section based signatures. The format is the same as for MD5 PE section based signatures. It can differentiate between them based on the length of the hash string in the signature. For best backwards compatibility, these should be placed inside a `*.msb` file. |
0 | 67 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,351 @@ |
0 |
+# Logical signatures |
|
1 |
+ |
|
2 |
+Logical signatures allow combining of multiple signatures in extended format using logical operators. They can provide both more detailed and flexible pattern matching. The logical sigs are stored inside `*.ldb` files in the following format: |
|
3 |
+ |
|
4 |
+``` |
|
5 |
+SignatureName;TargetDescriptionBlock;LogicalExpression;Subsig0; |
|
6 |
+Subsig1;Subsig2;... |
|
7 |
+``` |
|
8 |
+ |
|
9 |
+where: |
|
10 |
+ |
|
11 |
+- `TargetDescriptionBlock` provides information about the engine and target file with comma separated `Arg:Val` pairs. For args where `Val` is a range, the minimum and maximum values should be expressed as `min-max`. |
|
12 |
+ |
|
13 |
+- `LogicalExpression` specifies the logical expression describing the relationship between `Subsig0...SubsigN`. **Basis clause:** 0,1,...,N decimal indexes are SUB-EXPRESSIONS representing `Subsig0, Subsig1,...,SubsigN` respectively. **Inductive clause:** if `A` and `B` are SUB-EXPRESSIONS and `X, Y` are decimal numbers then `(A&B)`, `(A|B)`, `A=X`, `A=X,Y`, `A>X`, `A>X,Y`, `A<X` and `A<X,Y` are SUB-EXPRESSIONS |
|
14 |
+ |
|
15 |
+- `SubsigN` is n-th subsignature in extended format possibly preceded with an offset. There can be specified up to 64 subsigs. |
|
16 |
+ |
|
17 |
+Keywords used in `TargetDescriptionBlock`: |
|
18 |
+ |
|
19 |
+- `Target:X`: A number specifying the type of the target file: [Target Types](FileTypes.md#Target-Types). |
|
20 |
+ |
|
21 |
+- `Engine:X-Y`: Required engine functionality level (range; 0.96). Note that if the `Engine` keyword is used, it must be the first one in the `TargetDescriptionBlock` for backwards compatibility. See the [FLEVEL reference](FunctionalityLevels.md) for details. |
|
22 |
+ |
|
23 |
+- `FileSize:X-Y`: Required file size (range in bytes; 0.96) |
|
24 |
+ |
|
25 |
+- `EntryPoint`: Entry point offset (range in bytes; 0.96) |
|
26 |
+ |
|
27 |
+- `NumberOfSections`: Required number of sections in executable (range; 0.96) |
|
28 |
+ |
|
29 |
+- `Container:CL_TYPE_*`: File type of the container which stores the scanned file. |
|
30 |
+ |
|
31 |
+ Specifying `CL_TYPE_ANY` matches on root objects only (i.e. the target file is explicitely _not_ in a container). Chances slim that you would want to use `CL_TYPE_ANY` in a signature, because placing the malicious file in an archive will then prevent it from alerting. |
|
32 |
+ |
|
33 |
+ Every ClamAV file type has the potential to be a container for additional files, although some are more likely than others. When a file is parsed and data in the file is identified to be scanned as a unique type, that parent file becomes a container the moment the embedded content is scanned. For a list of possible CL_TYPEs, refer to the [File Types Reference](ClamAVFileTypes.md). |
|
34 |
+ |
|
35 |
+- `Intermediates:CL_TYPE_*>CL_TYPE_*`: Specify one or more layers of file types containing the scanned file. _This is an alternative to using `Container`._ |
|
36 |
+ |
|
37 |
+ You may specify up to 16 layers of file types separated by ’`>`’ in top-down order. Note that the ’`>`’ separator is not needed if you only specify a single container. The last type should be the immediate container containing the malicious file. Unlike with the `Container` option, `CL_TYPE_ANY` can be used as a wildcard file type. (expr; 0.100.0) |
|
38 |
+ |
|
39 |
+ For a list of possible CL_TYPEs, refer to the [File Types Reference](ClamAVFileTypes.md). |
|
40 |
+ |
|
41 |
+- `IconGroup1`: Icon group name 1 from .idb signature Required engine functionality (range; 0.96) |
|
42 |
+ |
|
43 |
+- `IconGroup2`: Icon group name 2 from .idb signature Required engine functionality (range; 0.96) |
|
44 |
+ |
|
45 |
+Modifiers for subexpressions: |
|
46 |
+ |
|
47 |
+- `A=X`: If the SUB-EXPRESSION A refers to a single signature then this signature must get matched exactly X times; if it refers to a (logical) block of signatures then this block must generate exactly X matches (with any of its sigs). |
|
48 |
+ |
|
49 |
+- `A=0` specifies negation (signature or block of signatures cannot be matched) |
|
50 |
+ |
|
51 |
+- `A=X,Y`: If the SUB-EXPRESSION A refers to a single signature then this signature must be matched exactly X times; if it refers to a (logical) block of signatures then this block must generate X matches and at least Y different signatures must get matched. |
|
52 |
+ |
|
53 |
+- `A>X`: If the SUB-EXPRESSION A refers to a single signature then this signature must get matched more than X times; if it refers to a (logical) block of signatures then this block must generate more than X matches (with any of its sigs). |
|
54 |
+ |
|
55 |
+- `A>X,Y`: If the SUB-EXPRESSION A refers to a single signature then this signature must get matched more than X times; if it refers to a (logical) block of signatures then this block must generate more than X matches _and_ at least Y different signatures must be matched. |
|
56 |
+ |
|
57 |
+- `A<X`: Just like `A>Z` above with the change of "more" to "less". |
|
58 |
+ |
|
59 |
+ If the SUB-EXPRESSION A refers to a single signature then this signature must get matched less than X times; if it refers to a (logical) block of signatures then this block must generate less than X matches (with any of its sigs). |
|
60 |
+ |
|
61 |
+- `A<X,Y`: Similar to `A>X,Y`. If the SUB-EXPRESSION A refers to a single signature then this signature must get matched less than X times; if it refers to a (logical) block of signatures then this block must generate less than X matches _and_ at least Y different signatures must be matched. |
|
62 |
+ |
|
63 |
+Examples: |
|
64 |
+ |
|
65 |
+``` |
|
66 |
+Sig1;Target:0;(0&1&2&3)&(4|1);6b6f74656b;616c61;7a6f6c77;7374656 |
|
67 |
+6616e;deadbeef |
|
68 |
+ |
|
69 |
+Sig2;Target:0;((0|1|2)>5,2)&(3|1);6b6f74656b;616c61;7a6f6c77;737 |
|
70 |
+46566616e |
|
71 |
+ |
|
72 |
+Sig3;Target:0;((0|1|2|3)=2)&(4|1);6b6f74656b;616c61;7a6f6c77;737 |
|
73 |
+46566616e;deadbeef |
|
74 |
+ |
|
75 |
+Sig4;Engine:51-255,Target:1;((0|1)&(2|3))&4;EP+123:33c06834f04100 |
|
76 |
+f2aef7d14951684cf04100e8110a00;S2+78:22??232c2d252229{-15}6e6573 |
|
77 |
+(63|64)61706528;S3+50:68efa311c3b9963cb1ee8e586d32aeb9043e;f9c58 |
|
78 |
+dcf43987e4f519d629b103375;SL+550:6300680065005c0046006900 |
|
79 |
+``` |
|
80 |
+ |
|
81 |
+## Subsignature Modifiers |
|
82 |
+ |
|
83 |
+ClamAV (clamav-0.99) supports a number of additional subsignature |
|
84 |
+modifiers for logical signatures. This is done by specifying `::` |
|
85 |
+followed by a number of characters representing the desired options. |
|
86 |
+Signatures using subsignature modifiers require `Engine:81-255` for |
|
87 |
+backwards-compatibility. |
|
88 |
+ |
|
89 |
+- Case-Insensitive \[`i`\] |
|
90 |
+ |
|
91 |
+ Specifying the `i` modifier causes ClamAV to match all alphabetic hex bytes as case-insensitive. All patterns in ClamAV are case-sensitive by default. |
|
92 |
+ |
|
93 |
+- Wide \[`w`\] |
|
94 |
+ |
|
95 |
+ Specifying the `w` causes ClamAV to match all hex bytes encoded with two bytes per character. Note this simply interweaves each character with NULL characters and does not truly support UTF-16 characters. Wildcards for ’wide’ subsignatures are not treated as wide (i.e. there can be an odd number of intermittent characters). This can be combined with `a` to search for patterns in both wide and ascii. |
|
96 |
+ |
|
97 |
+- Fullword \[`f`\] |
|
98 |
+ |
|
99 |
+ Match subsignature as a fullword (delimited by non-alphanumeric characters). |
|
100 |
+ |
|
101 |
+- Ascii \[`a`\] |
|
102 |
+ |
|
103 |
+ Match subsignature as ascii characters. This can be combined with `w` to search for patterns in both ascii and wide. |
|
104 |
+ |
|
105 |
+Examples: |
|
106 |
+ |
|
107 |
+``` |
|
108 |
+clamav-nocase-A;Engine:81-255,Target:0;0&1;41414141::i;424242424242::i |
|
109 |
+ -matches 'AAAA'(nocase) and 'BBBBBB'(nocase) |
|
110 |
+ |
|
111 |
+clamav-fullword-A;Engine:81-255,Target:0;0&1;414141;68656c6c6f::f |
|
112 |
+ -matches 'AAA' and 'hello'(fullword) |
|
113 |
+clamav-fullword-B;Engine:81-255,Target:0;0&1;414141;68656c6c6f::fi |
|
114 |
+ -matches 'AAA' and 'hello'(fullword nocase) |
|
115 |
+ |
|
116 |
+clamav-wide-B2;Engine:81-255,Target:0;0&1;414141;68656c6c6f::wa |
|
117 |
+ -matches 'AAA' and 'hello'(wide ascii) |
|
118 |
+clamav-wide-C0;Engine:81-255,Target:0;0&1;414141;68656c6c6f::iwfa |
|
119 |
+ -matches 'AAA' and 'hello'(nocase wide fullword ascii) |
|
120 |
+``` |
|
121 |
+ |
|
122 |
+## Special Subsignature Types |
|
123 |
+ |
|
124 |
+### Macro subsignatures |
|
125 |
+ |
|
126 |
+Introduced in ClamAV 0.96 |
|
127 |
+ |
|
128 |
+Format: `${min-max}MACROID$` |
|
129 |
+ |
|
130 |
+Macro subsignatures are used to combine a number of existing extended |
|
131 |
+signatures (`.ndb`) into a on-the-fly generated alternate string logical |
|
132 |
+signature (`.ldb`). Signatures using macro subsignatures require |
|
133 |
+`Engine:51-255` for backwards-compatibility. |
|
134 |
+ |
|
135 |
+Example: |
|
136 |
+ |
|
137 |
+``` |
|
138 |
+ test.ldb: |
|
139 |
+ TestMacro;Engine:51-255,Target:0;0&1;616161;${6-7}12$ |
|
140 |
+ |
|
141 |
+ test.ndb: |
|
142 |
+ D1:0:$12:626262 |
|
143 |
+ D2:0:$12:636363 |
|
144 |
+ D3:0:$30:626264 |
|
145 |
+``` |
|
146 |
+ |
|
147 |
+The example logical signature `TestMacro` is functionally equivalent |
|
148 |
+to: |
|
149 |
+ |
|
150 |
+``` |
|
151 |
+`TestMacro;Engine:51-255,Target:0;0;616161{3-4}(626262|636363)` |
|
152 |
+``` |
|
153 |
+ |
|
154 |
+- `MACROID` points to a group of signatures; there can be at most 32 macro groups. |
|
155 |
+ |
|
156 |
+ - In the example, `MACROID` is `12` and both `D1` and `D2` are members of macro group `12`. `D3` is a member of separate macro group `30`. |
|
157 |
+ |
|
158 |
+- `{min-max}` specifies the offset range at which one of the group signatures should match; the offset range is relative to the starting offset of the preceding subsignature. This means a macro subsignature cannot be the first subsignature. |
|
159 |
+ |
|
160 |
+ - In the example, `{min-max}` is `{6-7}` and it is relative to the start of a `616161` match. |
|
161 |
+ |
|
162 |
+- For more information and examples please see <https://bugzilla.clamav.net/show_bug.cgi?id=164>. |
|
163 |
+ |
|
164 |
+### Byte Compare Subsignatures |
|
165 |
+ |
|
166 |
+Introduced in ClamAV 0.101 |
|
167 |
+ |
|
168 |
+Format: `subsigid_trigger(offset#byte_options#comparisons)` |
|
169 |
+ |
|
170 |
+Byte compare subsignatures can be used to evaluate a numeric value at a given offset from the start of another (matched) subsignature within the same logical signature. These are executed after all other subsignatures within the logical subsignature are fired, with the exception of PCRE subsignatures. They can evaluate offsets only from a single referenced subsignature, and that subsignature must give a valid match for the evaluation to occur. |
|
171 |
+ |
|
172 |
+- `subsigid_trigger` is a required field and may refer to any single non-PCRE, non-Byte Compare subsignature within the lsig. The byte compare subsig will evaluate if `subsigid_trigger` matches. Triggering on multiple subsigs or logic based triggering is not currently supported. |
|
173 |
+ |
|
174 |
+- `offset` is a required field that consists of an `offset_modifier` and a numeric `offset` (hex or decimal offsets are okay). |
|
175 |
+ |
|
176 |
+ - `offset_modifier` can be either `>>` or `<<` where the former denotes a positive offset and the latter denotes a negative offset. The offset is calculated from the start of `subsigid_trigger`, which allows for byte extraction before the specified match, after the match, and within the match itself. |
|
177 |
+ |
|
178 |
+ - `offset` must be a positive hex or decimal value. This will be the number of bytes from the start of the referenced `subsigid_trigger` match within the file buffer to begin the comparison. |
|
179 |
+ |
|
180 |
+- `byte_options` are used to specify the numeric type and endianess of the extracted byte sequence in that order as well as the number of bytes to be read. By default ClamAV will attempt to matchup up to the number of byte specified, unless the `e` (exact) option is specified or the numeric type is `b` (binary). This field follows the form `[h|d|a|i][l|b][e]num_bytes` |
|
181 |
+ |
|
182 |
+ - `h|d|a|i` where `h` specifies the byte sequence will be in hex, `d` decimal, `a` automatic detection of hex or decimal at runtime, and `i` signifies raw binary data. |
|
183 |
+ |
|
184 |
+ - `l|b` where `l` specifies the byte sequence will be in little endian order and `b` big endian. If decimal `d` is specified, big-endian is implied and using `l` will result in a malformed database error. |
|
185 |
+ |
|
186 |
+ - `e` specifies that ClamAV will only evaluate the comparison if it can extract the exact number of bytes specified. This option is implicitly declared when using the `i` flag. |
|
187 |
+ |
|
188 |
+ - `num_bytes` specifies the number of bytes to extract. This can be a hex or decimal value. If `i` is specified only 1, 2, 4, and 8 are valid options. |
|
189 |
+ |
|
190 |
+- `comparisons` are a required field which denotes how to evaluate the extracted byte sequence. Each Byte Compare signature can have one or two `comparison_sets` separated by a comma. Each `comparison_set` consists of a `Comparison_symbol` and a `Comparison_value` and takes the form `Comparison_symbolComparison_value`. Thus, `comparisons` takes the form `comparison_set[,comparison_set]` |
|
191 |
+ |
|
192 |
+ - `Comparison_symbol` denotes the type of comparison to be done. The supported comparison symbols are `<`, `>`, `=`. |
|
193 |
+ |
|
194 |
+ - `Comparison_value` is a required field which must be a numeric hex or decimal value. If all other conditions are met, the byte compare subsig will evalutate the extracted byte sequence against this number based on the provided `comparison_symbol`. |
|
195 |
+ |
|
196 |
+### PCRE subsignatures |
|
197 |
+ |
|
198 |
+Introduced in ClamAV 0.99 |
|
199 |
+ |
|
200 |
+Format: `Trigger/PCRE/[Flags]` |
|
201 |
+ |
|
202 |
+PCRE subsignatures are used within a logical signature (`.ldb`) to specify regex matches that execute once triggered by a conditional based on preceding subsignatures. Signatures using PCRE subsignatures require `Engine:81-255` for backwards-compatibility. |
|
203 |
+ |
|
204 |
+- `Trigger` is a required field that is a valid `LogicalExpression` and may refer to any subsignatures that precede this subsignature. Triggers cannot be self-referential and cannot refer to subsequent subsignatures. |
|
205 |
+ |
|
206 |
+- `PCRE` is the expression representing the regex to execute. `PCRE` must be delimited by ’/’ and usage of ’/’ within the expression need to be escaped. For backward compatibility, ’;’ within the expression must be expressed as ’`\x3B`’. `PCRE` cannot be empty and (?UTF\*) control sequence is not allowed. If debug is specified, named capture groups are displayed in a post-execution report. |
|
207 |
+ |
|
208 |
+- `Flags` are a series of characters which affect the compilation and execution of `PCRE` within the PCRE compiler and the ClamAV engine. This field is optional. |
|
209 |
+ |
|
210 |
+ - `g [CLAMAV_GLOBAL]` specifies to search for ALL matches of PCRE (default is to search for first match). NOTE: INCREASES the time needed to run the PCRE. |
|
211 |
+ |
|
212 |
+ - `r [CLAMAV_ROLLING]` specifies to use the given offset as the starting location to search for a match as opposed to the only location; applies to subsigs without maxshifts. By default, in order to facilatate normal ClamAV offset behavior, PCREs are auto-anchored (only attempt match on first offset); using the rolling option disables the auto-anchoring. |
|
213 |
+ |
|
214 |
+ - `e [CLAMAV_ENCOMPASS]` specifies to CONFINE matching between the specified offset and maxshift; applies only when maxshift is specified. Note: DECREASES time needed to run the PCRE. |
|
215 |
+ |
|
216 |
+ - `i [PCRE_CASELESS]` |
|
217 |
+ |
|
218 |
+ - `s [PCRE_DOTALL]` |
|
219 |
+ |
|
220 |
+ - `m [PCRE_MULTILINE]` |
|
221 |
+ |
|
222 |
+ - `x [PCRE_EXTENDED]` |
|
223 |
+ |
|
224 |
+ - `A [PCRE_ANCHORED]` |
|
225 |
+ |
|
226 |
+ - `E [PCRE_DOLLAR_ENODNLY]` |
|
227 |
+ |
|
228 |
+ - `U [PCRE_UNGREEDY]` |
|
229 |
+ |
|
230 |
+Examples: |
|
231 |
+ |
|
232 |
+``` |
|
233 |
+Find.All.ClamAV;Engine:81-255,Target:0;1;6265676c6164697427736e6f7462797465636f6465;0/clamav/g |
|
234 |
+ |
|
235 |
+Find.ClamAV.OnlyAt.299;Engine:81-255,Target:0;2;7374756c747a67657473;7063726572656765786c6f6c;299:0&1/clamav/ |
|
236 |
+ |
|
237 |
+Find.ClamAV.StartAt.300;Engine:81-255,Target:0;3;616c61696e;62756731393238;636c6f736564;300:0&1&2/clamav/r |
|
238 |
+ |
|
239 |
+Find.All.Encompassed.ClamAV;Engine:81-255,Target:0;3;7768796172656e2774;796f757573696e67;79617261;200,300:0&1&2/clamav/ge |
|
240 |
+ |
|
241 |
+Named.CapGroup.Pcre;Engine:81-255,Target:0;3;636f75727479617264;616c62756d;74657272696572;50:0&1&2/variable=(?<nilshell>.{16})end/gr |
|
242 |
+ |
|
243 |
+Firefox.TreeRange.UseAfterFree;Engine:81-255,Target:0,Engine:81-255;0&1&2;2e766965772e73656c656374696f6e;2e696e76616c696461746553656c656374696f6e;0&1/\x2Eview\x2Eselection.*?\x2Etree\s*\x3D\s*null.*?\x2Einvalidate/smi |
|
244 |
+ |
|
245 |
+Firefox.IDB.UseAfterFree;Engine:81-255,Target:0;0&1;4944424b657952616e6765;0/^\x2e(only|lowerBound|upperBound|bound)\x28.*?\x29.*?\x2e(lower|upper|lowerOpen|upperOpen)/smi |
|
246 |
+ |
|
247 |
+Firefox.boundElements;Engine:81-255,Target:0;0&1&2;6576656e742e6 |
|
248 |
+26f756e64456c656d656e7473;77696e646f772e636c6f7365;0&1/on(load|click)\s*=\s*\x22?window\.close\s*\x28/si |
|
249 |
+``` |
|
250 |
+ |
|
251 |
+## Signatures for Version Information (VI) metadata in PE files |
|
252 |
+ |
|
253 |
+Starting with ClamAV 0.96 it is possible to easily match certain information built into PE files (executables and dynamic link libraries). Whenever you lookup the properties of a PE executable file in windows, you are presented with a bunch of details about the file itself. |
|
254 |
+ |
|
255 |
+These info are stored in a special area of the file resources which goes under the name of `VS_VERSION_INFORMATION` (or versioninfo for short). It is divided into 2 parts. The first part (which is rather uninteresting) is really a bunch of numbers and flags indicating the product and file version. It was originally intended for use with installers which, after parsing it, should be able to determine whether a certain executable or library are to be upgraded/overwritten or are already up to date. Suffice to say, this approach never really worked and is generally never used. |
|
256 |
+ |
|
257 |
+The second block is much more interesting: it is a simple list of key/value strings, intended for user information and completely ignored by the OS. For example, if you look at ping.exe you can see the company being *"Microsoft Corporation"*, the description *"TCP/IP Ping command"*, the internal name *"ping.exe"* and so on... Depending on the OS version, some keys may be given peculiar visibility in the file properties dialog, however they are internally all the same. |
|
258 |
+ |
|
259 |
+To match a versioninfo key/value pair, the special file offset anchor `VI` was introduced. This is similar to the other anchors (like `EP` and `SL`) except that, instead of matching the hex pattern against a single offset, it checks it against each and every key/value pair in the file. The `VI` token doesn’t need nor accept a `+/-` offset like e.g. `EP+1`. As for the hex signature itself, it’s just the utf16 dump of the key and value. Only the `??` and `(aa|bb)` wildcards are allowed in the signature. Usually, you don’t need to bother figuring it out: each key/value pair together with the corresponding VI-based signature is printed by `clamscan` when the `--debug` option is given. |
|
260 |
+ |
|
261 |
+For example `clamscan --debug freecell.exe` produces: |
|
262 |
+ |
|
263 |
+```bash |
|
264 |
+[...] |
|
265 |
+Recognized MS-EXE/DLL file |
|
266 |
+in cli_peheader |
|
267 |
+versioninfo_cb: type: 10, name: 1, lang: 410, rva: 9608 |
|
268 |
+cli_peheader: parsing version info @ rva 9608 (1/1) |
|
269 |
+VersionInfo (d2de): 'CompanyName'='Microsoft Corporation' - |
|
270 |
+VI:43006f006d00700061006e0079004e0061006d006500000000004d006900 |
|
271 |
+630072006f0073006f0066007400200043006f00720070006f0072006100740 |
|
272 |
+069006f006e000000 |
|
273 |
+VersionInfo (d32a): 'FileDescription'='Entertainment Pack |
|
274 |
+FreeCell Game' - VI:460069006c006500440065007300630072006900700 |
|
275 |
+0740069006f006e000000000045006e007400650072007400610069006e006d |
|
276 |
+0065006e00740020005000610063006b0020004600720065006500430065006 |
|
277 |
+c006c002000470061006d0065000000 |
|
278 |
+VersionInfo (d396): 'FileVersion'='5.1.2600.0 (xpclient.010817 |
|
279 |
+-1148)' - VI:460069006c006500560065007200730069006f006e00000000 |
|
280 |
+0035002e0031002e0032003600300030002e003000200028007800700063006 |
|
281 |
+c00690065006e0074002e003000310030003800310037002d00310031003400 |
|
282 |
+380029000000 |
|
283 |
+VersionInfo (d3fa): 'InternalName'='freecell' - VI:49006e007400 |
|
284 |
+650072006e0061006c004e0061006d006500000066007200650065006300650 |
|
285 |
+06c006c000000 |
|
286 |
+VersionInfo (d4ba): 'OriginalFilename'='freecell' - VI:4f007200 |
|
287 |
+6900670069006e0061006c00460069006c0065006e0061006d0065000000660 |
|
288 |
+0720065006500630065006c006c000000 |
|
289 |
+VersionInfo (d4f6): 'ProductName'='Sistema operativo Microsoft |
|
290 |
+Windows' - VI:500072006f0064007500630074004e0061006d00650000000 |
|
291 |
+000530069007300740065006d00610020006f00700065007200610074006900 |
|
292 |
+76006f0020004d006900630072006f0073006f0066007400ae0020005700690 |
|
293 |
+06e0064006f0077007300ae000000 |
|
294 |
+VersionInfo (d562): 'ProductVersion'='5.1.2600.0' - VI:50007200 |
|
295 |
+6f006400750063007400560065007200730069006f006e00000035002e00310 |
|
296 |
+02e0032003600300030002e0030000000 |
|
297 |
+[...] |
|
298 |
+``` |
|
299 |
+ |
|
300 |
+Although VI-based signatures are intended for use in logical signatures you can test them using ordinary `.ndb` files. For example: |
|
301 |
+ |
|
302 |
+``` |
|
303 |
+ my_test_vi_sig:1:VI:paste_your_hex_sig_here |
|
304 |
+``` |
|
305 |
+ |
|
306 |
+Final note. If you want to decode a VI-based signature into a human readable form you can use: |
|
307 |
+ |
|
308 |
+```bash |
|
309 |
+echo hex_string | xxd -r -p | strings -el |
|
310 |
+``` |
|
311 |
+ |
|
312 |
+For example: |
|
313 |
+ |
|
314 |
+```bash |
|
315 |
+$ echo 460069006c0065004400650073006300720069007000740069006f006e |
|
316 |
+000000000045006e007400650072007400610069006e006d0065006e007400200 |
|
317 |
+05000610063006b0020004600720065006500430065006c006c00200047006100 |
|
318 |
+6d0065000000 | xxd -r -p | strings -el |
|
319 |
+FileDescription |
|
320 |
+Entertainment Pack FreeCell Game |
|
321 |
+``` |
|
322 |
+ |
|
323 |
+## Icon Signatures for PE files |
|
324 |
+ |
|
325 |
+While Icon Signatures are stored in a `.idb` file, they are a feature of Logical Signatures. |
|
326 |
+ |
|
327 |
+ClamAV 0.96 includes an approximate/fuzzy icon matcher to help detecting malicious executables disguising themselves as innocent looking image files, office documents and the like. |
|
328 |
+ |
|
329 |
+Icon matching is only triggered by Logical Signatures (`.ldb`) using the special attribute tokens `IconGroup1` or `IconGroup2`. These identify two (optional) groups of icons defined in a `.idb` database file. The format of the `.idb` file is: |
|
330 |
+ |
|
331 |
+``` |
|
332 |
+ ICONNAME:GROUP1:GROUP2:ICON_HASH |
|
333 |
+``` |
|
334 |
+ |
|
335 |
+where: |
|
336 |
+ |
|
337 |
+- `ICON_NAME` is a unique string identifier for a specific icon, |
|
338 |
+ |
|
339 |
+- `GROUP1` is a string identifier for the first group of icons (`IconGroup1`) |
|
340 |
+ |
|
341 |
+- `GROUP2` is a string identifier for the second group of icons (`IconGroup2`), |
|
342 |
+ |
|
343 |
+- `ICON_HASH` is a fuzzy hash of the icon image |
|
344 |
+ |
|
345 |
+The `ICON_HASH` field can be obtained from the debug output of libclamav. For example: |
|
346 |
+ |
|
347 |
+```bash |
|
348 |
+LibClamAV debug: ICO SIGNATURE: |
|
349 |
+ICON_NAME:GROUP1:GROUP2:18e2e0304ce60a0cc3a09053a30000414100057e000afe0000e 80006e510078b0a08910d11ad04105e0811510f084e01040c080a1d0b0021000a39002a41 |
|
350 |
+``` |
0 | 351 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,682 @@ |
0 |
+# PhishSigs |
|
1 |
+ |
|
2 |
+Table of Contents |
|
3 |
+- [PhishSigs](#phishsigs) |
|
4 |
+- [Database file format](#database-file-format) |
|
5 |
+ - [PDB format](#pdb-format) |
|
6 |
+ - [GDB format](#gdb-format) |
|
7 |
+ - [WDB format](#wdb-format) |
|
8 |
+ - [Hints](#hints) |
|
9 |
+ - [Examples of PDB signatures](#examples-of-pdb-signatures) |
|
10 |
+ - [Examples of WDB signatures](#examples-of-wdb-signatures) |
|
11 |
+ - [Example for how the URL extractor works](#example-for-how-the-url-extractor-works) |
|
12 |
+ - [How matching works](#how-matching-works) |
|
13 |
+ - [RealURL, displayedURL concatenation](#realurl-displayedurl-concatenation) |
|
14 |
+ - [What happens when a match is found](#what-happens-when-a-match-is-found) |
|
15 |
+ - [Extraction of realURL, displayedURL from HTML tags](#extraction-of-realurl-displayedurl-from-html-tags) |
|
16 |
+ - [Example](#example) |
|
17 |
+ - [Simple patterns](#simple-patterns) |
|
18 |
+ - [Regular expressions](#regular-expressions) |
|
19 |
+ - [Flags](#flags) |
|
20 |
+- [Introduction to regular expressions](#introduction-to-regular-expressions) |
|
21 |
+ - [Special characters](#special-characters) |
|
22 |
+ - [Character classes](#character-classes) |
|
23 |
+ - [Escaping](#escaping) |
|
24 |
+ - [Alternation](#alternation) |
|
25 |
+ - [Optional matching, and repetition](#optional-matching-and-repetition) |
|
26 |
+ - [Groups](#groups) |
|
27 |
+- [How to create database files](#how-to-create-database-files) |
|
28 |
+ - [How to create and maintain the whitelist (daily.wdb)](#how-to-create-and-maintain-the-whitelist-dailywdb) |
|
29 |
+ - [How to create and maintain the domainlist (daily.pdb)](#how-to-create-and-maintain-the-domainlist-dailypdb) |
|
30 |
+ - [Dealing with false positives, and undetected phishing mails](#dealing-with-false-positives-and-undetected-phishing-mails) |
|
31 |
+ - [False positives](#false-positives) |
|
32 |
+ - [Undetected phish mails](#undetected-phish-mails) |
|
33 |
+ |
|
34 |
+# Database file format |
|
35 |
+ |
|
36 |
+## PDB format |
|
37 |
+ |
|
38 |
+This file contains urls/hosts that are target of phishing attempts. It |
|
39 |
+contains lines in the following format: |
|
40 |
+ |
|
41 |
+``` |
|
42 |
+ R[Filter]:RealURL:DisplayedURL[:FuncLevelSpec] |
|
43 |
+ H[Filter]:DisplayedHostname[:FuncLevelSpec] |
|
44 |
+``` |
|
45 |
+ |
|
46 |
+- `R` |
|
47 |
+ |
|
48 |
+ regular expression, for the concatenated URL |
|
49 |
+ |
|
50 |
+- `H` |
|
51 |
+ |
|
52 |
+ matches the `DisplayedHostname` as a simple pattern (literally, no regular expression) |
|
53 |
+ |
|
54 |
+ - the pattern can match either the full hostname |
|
55 |
+ |
|
56 |
+ - or a subdomain of the specified hostname |
|
57 |
+ |
|
58 |
+ - to avoid false matches in case of subdomain matches, the engine checks that there is a dot(`.`) or a space(` `) before the matched portion |
|
59 |
+ |
|
60 |
+- `Filter` |
|
61 |
+ |
|
62 |
+ is ignored for R and H for compatibility reasons |
|
63 |
+ |
|
64 |
+- `RealURL` |
|
65 |
+ |
|
66 |
+ is the URL the user is sent to, example: *href* attribute of an html anchor (*\<a\> tag*) |
|
67 |
+ |
|
68 |
+- `DisplayedURL` |
|
69 |
+ |
|
70 |
+ is the URL description displayed to the user, where its *claimed* they are sent, example: contents of an html anchor (*\<a\> tag*) |
|
71 |
+ |
|
72 |
+- `DisplayedHostname` |
|
73 |
+ |
|
74 |
+ is the hostname portion of the DisplayedURL |
|
75 |
+ |
|
76 |
+- `FuncLevelSpec` |
|
77 |
+ |
|
78 |
+ an (optional) functionality level, 2 formats are possible: |
|
79 |
+ |
|
80 |
+ - `minlevel` all engines having functionality level \>= `minlevel` will load this line |
|
81 |
+ |
|
82 |
+ - `minlevel-maxlevel` engines with functionality level \(>=\) `minlevel`, and \(<\) `maxlevel` will load this line |
|
83 |
+ |
|
84 |
+## GDB format |
|
85 |
+ |
|
86 |
+This file contains URL hashes in the following format: |
|
87 |
+ |
|
88 |
+ S:P:HostPrefix[:FuncLevelSpec] |
|
89 |
+ S:F:Sha256hash[:FuncLevelSpec] |
|
90 |
+ S1:P:HostPrefix[:FuncLevelSpec] |
|
91 |
+ S1:F:Sha256hash[:FuncLevelSpec] |
|
92 |
+ S2:P:HostPrefix[:FuncLevelSpec] |
|
93 |
+ S2:F:Sha256hash[:FuncLevelSpec] |
|
94 |
+ S:W:Sha256hash[:FuncLevelSpec] |
|
95 |
+ |
|
96 |
+- `S:` |
|
97 |
+ |
|
98 |
+ These are hashes for Google Safe Browsing - malware sites, and should not be used for other purposes. |
|
99 |
+ |
|
100 |
+- `S2:` |
|
101 |
+ |
|
102 |
+ These are hashes for Google Safe Browsing - phishing sites, and should not be used for other purposes. |
|
103 |
+ |
|
104 |
+- `S1:` |
|
105 |
+ |
|
106 |
+ Hashes for blacklisting phishing sites. Virus name: Phishing.URL.Blacklisted |
|
107 |
+ |
|
108 |
+- `S:W:` |
|
109 |
+ |
|
110 |
+ Locally whitelisted hashes. |
|
111 |
+ |
|
112 |
+- `HostPrefix` |
|
113 |
+ |
|
114 |
+ 4-byte prefix of the sha256 hash of the last 2 or 3 components of the hostname. If prefix doesn’t match, no further lookups are performed. |
|
115 |
+ |
|
116 |
+- `Sha256hash` |
|
117 |
+ |
|
118 |
+ sha256 hash of the canonicalized URL, or a sha256 hash of its prefix/suffix according to the Google Safe Browsing “Performing Lookups” rules. There should be a corresponding `:P:HostkeyPrefix` entry for the hash to be taken into consideration. |
|
119 |
+ |
|
120 |
+To see which hash/URL matched, look at the `clamscan --debug` output, and look for the following strings: `Looking up hash`, `prefix matched`, and `Hash matched`. Local whitelisting of .gdb entries can be done by creating a local.gdb file, and adding a line `S:W:<HASH>`. |
|
121 |
+ |
|
122 |
+## WDB format |
|
123 |
+ |
|
124 |
+This file contains whitelisted url pairs It contains lines in the following format: |
|
125 |
+ |
|
126 |
+``` |
|
127 |
+ X:RealURL:DisplayedURL[:FuncLevelSpec] |
|
128 |
+ M:RealHostname:DisplayedHostname[:FuncLevelSpec] |
|
129 |
+``` |
|
130 |
+ |
|
131 |
+- `X` |
|
132 |
+ |
|
133 |
+ regular expression, for the *entire URL*, not just the hostname |
|
134 |
+ |
|
135 |
+ - The regular expression is by default anchored to start-of-line and end-of-line, as if you have used `^RegularExpression$` |
|
136 |
+ |
|
137 |
+ - A trailing `/` is automatically added both to the regex, and the input string to avoid false matches |
|
138 |
+ |
|
139 |
+ - The regular expression matches the *concatenation* of the RealURL, a colon(`:`), and the DisplayedURL as a single string. It doesn’t separately match RealURL and DisplayedURL\! |
|
140 |
+ |
|
141 |
+- `M` |
|
142 |
+ |
|
143 |
+ matches hostname, or subdomain of it, see notes for H above |
|
144 |
+ |
|
145 |
+## Hints |
|
146 |
+ |
|
147 |
+- empty lines are ignored |
|
148 |
+ |
|
149 |
+- the colons are mandatory |
|
150 |
+ |
|
151 |
+- Don’t leave extra spaces on the end of a line\! |
|
152 |
+ |
|
153 |
+- if any of the lines don’t conform to this format, clamav will abort with a Malformed Database Error |
|
154 |
+ |
|
155 |
+- see section [Extraction-of-realURL](#Extraction-of-realURL,-displayedURL-from-HTML-tags) for more details on realURL/displayedURL |
|
156 |
+ |
|
157 |
+## Examples of PDB signatures |
|
158 |
+ |
|
159 |
+To check for phishing mails that target amazon.com, or subdomains of |
|
160 |
+amazon.com: |
|
161 |
+ |
|
162 |
+``` |
|
163 |
+ H:amazon.com |
|
164 |
+``` |
|
165 |
+ |
|
166 |
+To do the same, but for amazon.co.uk: |
|
167 |
+ |
|
168 |
+``` |
|
169 |
+ H:amazon.co.uk |
|
170 |
+``` |
|
171 |
+ |
|
172 |
+To limit the signatures to certain engine versions: |
|
173 |
+ |
|
174 |
+``` |
|
175 |
+ H:amazon.co.uk:20-30 |
|
176 |
+ H:amazon.co.uk:20- |
|
177 |
+ H:amazon.co.uk:0-20 |
|
178 |
+``` |
|
179 |
+ |
|
180 |
+First line: engine versions 20, 21, ..., 29 can load it |
|
181 |
+ |
|
182 |
+Second line: engine versions \>= 20 can load it |
|
183 |
+ |
|
184 |
+Third line: engine versions \< 20 can load it |
|
185 |
+ |
|
186 |
+In a real situation, you’d probably use the second form. A situation like that would be if you are using a feature of the signatures not available in earlier versions, or if earlier versions have bugs with your signature. Its neither case here, the above examples are for illustrative purposes only. |
|
187 |
+ |
|
188 |
+## Examples of WDB signatures |
|
189 |
+ |
|
190 |
+To allow amazon’s country specific domains and amazon.com, to mix domain names in DisplayedURL, and RealURL: |
|
191 |
+ |
|
192 |
+ X:.+\.amazon\.(at|ca|co\.uk|co\.jp|de|fr)([/?].*)?:.+\.amazon\.com([/?].*)?:17- |
|
193 |
+ |
|
194 |
+Explanation of this signature: |
|
195 |
+ |
|
196 |
+- `X:` |
|
197 |
+ |
|
198 |
+ this is a regular expression |
|
199 |
+ |
|
200 |
+- `:17-` |
|
201 |
+ |
|
202 |
+ load signature only for engines with functionality level \>= 17 (recommended for type X) |
|
203 |
+ |
|
204 |
+The regular expression is the following (X:, :17- stripped, and a / appended) |
|
205 |
+ |
|
206 |
+``` |
|
207 |
+ .+\.amazon\.(at|ca|co\.uk|co\.jp|de|fr)([/?].*)?:.+\.amazon\.com([/?].*)?/ |
|
208 |
+``` |
|
209 |
+ |
|
210 |
+Explanation of this regular expression (note that it is a single regular expression, and not 2 regular expressions splitted at the :). |
|
211 |
+ |
|
212 |
+- `.+` |
|
213 |
+ |
|
214 |
+ any subdomain of |
|
215 |
+ |
|
216 |
+- `\.amazon\.` |
|
217 |
+ |
|
218 |
+ domain we are whitelisting (RealURL part) |
|
219 |
+ |
|
220 |
+- `(at|ca|co\.uk|co\.jp|de|fr)` |
|
221 |
+ |
|
222 |
+ country-domains: at, ca, co.uk, co.jp, de, fr |
|
223 |
+ |
|
224 |
+- `([/?].*)?` |
|
225 |
+ |
|
226 |
+ recomended way to end real url part of whitelist, this protects against embedded URLs (evilurl.example.com/amazon.co.uk/) |
|
227 |
+ |
|
228 |
+- `:` |
|
229 |
+ |
|
230 |
+ RealURL and DisplayedURL are concatenated via a :, so match a literal : here |
|
231 |
+ |
|
232 |
+- `.+` |
|
233 |
+ |
|
234 |
+ any subdomain of |
|
235 |
+ |
|
236 |
+- `\.amazon\.com` |
|
237 |
+ |
|
238 |
+ whitelisted DisplayedURL |
|
239 |
+ |
|
240 |
+- `([/?].*)?` |
|
241 |
+ |
|
242 |
+ recommended way to end displayed url part, to protect against embedded URLs |
|
243 |
+ |
|
244 |
+- `/` |
|
245 |
+ |
|
246 |
+ automatically added to further protect against embedded URLs |
|
247 |
+ |
|
248 |
+When you whitelist an entry make sure you check that both domains are owned by the same entity. What this whitelist entry allows is: Links claiming to point to amazon.com (DisplayedURL), but really go to country-specific domain of amazon (RealURL). |
|
249 |
+ |
|
250 |
+## Example for how the URL extractor works |
|
251 |
+ |
|
252 |
+Consider the following HTML file: |
|
253 |
+ |
|
254 |
+```html |
|
255 |
+ <html> |
|
256 |
+ <a href="http://1.realurl.example.com/"> |
|
257 |
+ 1.displayedurl.example.com |
|
258 |
+ </a> |
|
259 |
+ <a href="http://2.realurl.example.com"> |
|
260 |
+ 2 d<b>i<p>splayedurl.e</b>xa<i>mple.com |
|
261 |
+ </a> |
|
262 |
+ <a href="http://3.realurl.example.com"> |
|
263 |
+ 3.nested.example.com |
|
264 |
+ <a href="http://4.realurl.example.com"> |
|
265 |
+ 4.displayedurl.example.com |
|
266 |
+ </a> |
|
267 |
+ </a> |
|
268 |
+ <form action="http://5.realurl.example.com"> |
|
269 |
+ sometext |
|
270 |
+ <img src="http://5.displayedurl.example.com/img0.gif"/> |
|
271 |
+ <a href="http://5.form.nested.displayedurl.example.com"> |
|
272 |
+ 5.form.nested.link-displayedurl.example.com |
|
273 |
+ </a> |
|
274 |
+ </form> |
|
275 |
+ <a href="http://6.realurl.example.com"> |
|
276 |
+ 6.displ |
|
277 |
+ <img src="6.displayedurl.example.com/img1.gif"/> |
|
278 |
+ ayedurl.example.com |
|
279 |
+ </a> |
|
280 |
+ <a href="http://7.realurl.example.com"> |
|
281 |
+ <iframe src="http://7.displayedurl.example.com"> |
|
282 |
+ </a> |
|
283 |
+``` |
|
284 |
+ |
|
285 |
+The phishing engine extract the following |
|
286 |
+RealURL/DisplayedURL pairs from it: |
|
287 |
+ |
|
288 |
+``` |
|
289 |
+ http://1.realurl.example.com/ |
|
290 |
+ 1.displayedurl.example.com |
|
291 |
+ |
|
292 |
+ http://2.realurl.example.com |
|
293 |
+ 2displayedurl.example.com |
|
294 |
+ |
|
295 |
+ http://3.realurl.example.com |
|
296 |
+ 3.nested.example.com |
|
297 |
+ |
|
298 |
+ http://4.realurl.example.com |
|
299 |
+ 4.displayedurl.example.com |
|
300 |
+ |
|
301 |
+ http://5.realurl.example.com |
|
302 |
+ http://5.displayedurl.example.com/img0.gif |
|
303 |
+ |
|
304 |
+ http://5.realurl.example.com |
|
305 |
+ http://5.form.nested.displayedurl.example.com |
|
306 |
+ |
|
307 |
+ http://5.form.nested.displayedurl.example.com |
|
308 |
+ 5.form.nested.link-displayedurl.example.com |
|
309 |
+ |
|
310 |
+ http://6.realurl.example.com |
|
311 |
+ 6.displayedurl.example.com |
|
312 |
+ |
|
313 |
+ http://6.realurl.example.com |
|
314 |
+ 6.displayedurl.example.com/img1.gif |
|
315 |
+``` |
|
316 |
+ |
|
317 |
+## How matching works |
|
318 |
+ |
|
319 |
+### RealURL, displayedURL concatenation |
|
320 |
+ |
|
321 |
+The phishing detection module processes pairs of RealURL/DisplayedURL. Matching against daily.wdb is done as follows: the realURL is concatenated with a `:`, and with the DisplayedURL, then that *line* is matched against the lines in daily.wdb/daily.pdb |
|
322 |
+ |
|
323 |
+So if you have this line in daily.wdb: |
|
324 |
+ |
|
325 |
+ M:www.google.ro:www.google.com |
|
326 |
+ |
|
327 |
+and this href: `<a href='http://www.google.ro'>www.google.com</a>` then it will be whitelisted, but: `<a href='http://images.google.com'>www.google.com</a>` will not. |
|
328 |
+ |
|
329 |
+### What happens when a match is found |
|
330 |
+ |
|
331 |
+In the case of the whitelist, a match means that the RealURL/DisplayedURL combination is considered clean, and no further checks are performed on it. |
|
332 |
+ |
|
333 |
+In the case of the domainlist, a match means that the RealURL/displayedURL is going to be checked for phishing attempts. |
|
334 |
+ |
|
335 |
+Furthermore you can restrict what checks are to be performed by specifying the 3-digit hexnumber. |
|
336 |
+ |
|
337 |
+### Extraction of realURL, displayedURL from HTML tags |
|
338 |
+ |
|
339 |
+The html parser extracts pairs of realURL/displayedURL based on the following rules. |
|
340 |
+ |
|
341 |
+In version 0.93: After URLs have been extracted, they are normalized, and cut after the hostname. `http://test.example.com/path/somecgi?queryparameters` becomes `http://test.example.com/` |
|
342 |
+ |
|
343 |
+- `a` |
|
344 |
+ |
|
345 |
+ (anchor) the *href* is the realURL, its *contents* is the displayedURL |
|
346 |
+ |
|
347 |
+ - contents |
|
348 |
+ is the tag-stripped contents of the \<a\> tags, so for example \<b\> tags are stripped (but not their contents) |
|
349 |
+ |
|
350 |
+ nesting another \<a\> tag withing an \<a\> tag (besides being invalid html) is treated as a \</a\>\<a.. |
|
351 |
+ |
|
352 |
+- `form` |
|
353 |
+ |
|
354 |
+ the *action* attribute is the realURL, and a nested \<a\> tag is the displayedURL |
|
355 |
+ |
|
356 |
+- `img/area` |
|
357 |
+ |
|
358 |
+ if nested within an *\<a\>* tag, the realURL is the *href* of the a tag, and the *src/dynsrc/area* is the displayedURL of the img |
|
359 |
+ |
|
360 |
+ if nested withing a *form* tag, then the action attribute of the *form* tag is the realURL |
|
361 |
+ |
|
362 |
+- `iframe` |
|
363 |
+ |
|
364 |
+ if nested withing an *\<a\>* tag the *src* attribute is the displayedURL, and the *href* of its parent *a* tag is the realURL |
|
365 |
+ |
|
366 |
+ if nested withing a *form* tag, then the action attribute of the *form* tag is the realURL |
|
367 |
+ |
|
368 |
+### Example |
|
369 |
+ |
|
370 |
+Consider this html file: |
|
371 |
+ |
|
372 |
+```html |
|
373 |
+<a href=”evilurl”\>www.paypal.com\</a\>* |
|
374 |
+ |
|
375 |
+<a href=”evilurl2” title=”www.ebay.com”\>click here to sign |
|
376 |
+in\</a\>* |
|
377 |
+ |
|
378 |
+<form action=”evilurl_form”\>* |
|
379 |
+ |
|
380 |
+*Please sign in to \<a href=”cgi.ebay.com”\>Ebay\</a\using this |
|
381 |
+form* |
|
382 |
+ |
|
383 |
+<input type=’text’ name=’username’\>Username\</input\>* |
|
384 |
+ |
|
385 |
+*....* |
|
386 |
+ |
|
387 |
+</form\>* |
|
388 |
+ |
|
389 |
+<a href=”evilurl”\>\<img src=”images.paypal.com/secure.jpg”\>\</a\>* |
|
390 |
+``` |
|
391 |
+ |
|
392 |
+The resulting realURL/displayedURL pairs will be (note that one tag can generate multiple pairs): |
|
393 |
+ |
|
394 |
+- evilurl / www.paypal.com |
|
395 |
+ |
|
396 |
+- evilurl2 / click here to sign in |
|
397 |
+ |
|
398 |
+- evilurl2 / www.ebay.com |
|
399 |
+ |
|
400 |
+- evilurl_form / cgi.ebay.com |
|
401 |
+ |
|
402 |
+- cgi.ebay.com / Ebay |
|
403 |
+ |
|
404 |
+- evilurl / image.paypal.com/secure.jpg |
|
405 |
+ |
|
406 |
+## Simple patterns |
|
407 |
+ |
|
408 |
+Simple patterns are matched literally, i.e. if you say: |
|
409 |
+ |
|
410 |
+``` |
|
411 |
+www.google.com |
|
412 |
+``` |
|
413 |
+ |
|
414 |
+it is going to match *www.google.com*, and only that. The *. (dot)* character has no special meaning (see the section on regexes [\[sec:Regular-expressions\]](#sec:Regular-expressions) for how the *.(dot)* character behaves there) |
|
415 |
+ |
|
416 |
+## Regular expressions |
|
417 |
+ |
|
418 |
+POSIX regular expressions are supported, and you can consider that internally it is wrapped by *^*, and *$.* In other words, this means that the regular expression has to match the entire concatenated (see section [RealURL,-displayedURL-concatenation](#RealURL,-displayedURL-concatenation) for details on concatenation) url. |
|
419 |
+ |
|
420 |
+It is recomended that you read section [Introduction-to-regular](#Introduction-to-regular) to learn how to write regular expressions, and then come back and read this for hints. |
|
421 |
+ |
|
422 |
+Be advised that clamav contains an internal, very basic regex matcher to reduce the load on the regex matching core. Thus it is recomended that you avoid using regex syntax not supported by it at the very beginning of regexes (at least the first few characters). |
|
423 |
+ |
|
424 |
+Currently the clamav regex matcher supports: |
|
425 |
+ |
|
426 |
+- `.` (dot) character |
|
427 |
+ |
|
428 |
+- `\(\backslash\)` (escaping special characters) |
|
429 |
+ |
|
430 |
+- `|` (pipe) alternatives |
|
431 |
+ |
|
432 |
+- `\[\]` (character classes) |
|
433 |
+ |
|
434 |
+- `()` (parenthesis for grouping, but no group extraction is performed) |
|
435 |
+ |
|
436 |
+- other non-special characters |
|
437 |
+ |
|
438 |
+Thus the following are not supported: |
|
439 |
+ |
|
440 |
+- `\+` repetition |
|
441 |
+ |
|
442 |
+- `\*` repetition |
|
443 |
+ |
|
444 |
+- `{}` repetition |
|
445 |
+ |
|
446 |
+- backreferences |
|
447 |
+ |
|
448 |
+- lookaround |
|
449 |
+ |
|
450 |
+- other “advanced” features not listed in the supported list ;) |
|
451 |
+ |
|
452 |
+This however shouldn’t discourage you from using the “not directly supported features “, because if the internal engine encounters unsupported syntax, it passes it on to the POSIX regex core (beginning from the first unsupported token, everything before that is still processed by the internal matcher). An example might make this more clear: |
|
453 |
+ |
|
454 |
+*www\(\backslash\).google\(\backslash\).(com|ro|it) (\[a-zA-Z\])+\(\backslash\).google\(\backslash\).(com|ro|it)* |
|
455 |
+ |
|
456 |
+Everything till *(\[a-zA-Z\])+* is processed internally, that parenthesis (and everything beyond) is processed by the posix core. |
|
457 |
+ |
|
458 |
+Examples of url pairs that match: |
|
459 |
+ |
|
460 |
+- *www.google.ro images.google.ro* |
|
461 |
+ |
|
462 |
+- www.google.com images.google.ro |
|
463 |
+ |
|
464 |
+Example of url pairs that don’t match: |
|
465 |
+ |
|
466 |
+- www.google.ro images1.google.ro |
|
467 |
+ |
|
468 |
+- images.google.com image.google.com |
|
469 |
+ |
|
470 |
+## Flags |
|
471 |
+ |
|
472 |
+Flags are a binary OR of the following numbers: |
|
473 |
+ |
|
474 |
+- HOST_SUFFICIENT |
|
475 |
+ |
|
476 |
+ 1 |
|
477 |
+ |
|
478 |
+- DOMAIN_SUFFICIENT |
|
479 |
+ |
|
480 |
+ 2 |
|
481 |
+ |
|
482 |
+- DO_REVERSE_LOOKUP |
|
483 |
+ |
|
484 |
+ 4 |
|
485 |
+ |
|
486 |
+- CHECK_REDIR |
|
487 |
+ |
|
488 |
+ 8 |
|
489 |
+ |
|
490 |
+- CHECK_SSL |
|
491 |
+ |
|
492 |
+ 16 |
|
493 |
+ |
|
494 |
+- CHECK_CLOAKING |
|
495 |
+ |
|
496 |
+ 32 |
|
497 |
+ |
|
498 |
+- CLEANUP_URL |
|
499 |
+ |
|
500 |
+ 64 |
|
501 |
+ |
|
502 |
+- CHECK_DOMAIN_REVERSE |
|
503 |
+ |
|
504 |
+ 128 |
|
505 |
+ |
|
506 |
+- CHECK_IMG_URL |
|
507 |
+ |
|
508 |
+ 256 |
|
509 |
+ |
|
510 |
+- DOMAINLIST_REQUIRED |
|
511 |
+ |
|
512 |
+ 512 |
|
513 |
+ |
|
514 |
+The names of the constants are self-explanatory. |
|
515 |
+ |
|
516 |
+These constants are defined in libclamav/phishcheck.h, you can check there for the latest flags. |
|
517 |
+ |
|
518 |
+There is a default set of flags that are enabled, these are currently: |
|
519 |
+ |
|
520 |
+ ( CLEANUP_URL | CHECK_SSL | CHECK_CLOAKING | CHECK_IMG_URL ) |
|
521 |
+ |
|
522 |
+ssl checking is performed only for a tags currently. |
|
523 |
+ |
|
524 |
+You must decide for each line in the domainlist if you want to filter any flags (that is you don’t want certain checks to be done), and then calculate the binary OR of those constants, and then convert it into a 3-digit hexnumber. For example you devide that domain_sufficient shouldn’t be used for ebay.com, and you don’t want to check images either, so you come up with this flag number: \(2|256\Rightarrow\)258\((decimal)\Rightarrow102(hexadecimal)\) |
|
525 |
+ |
|
526 |
+So you add this line to daily.wdb: |
|
527 |
+ |
|
528 |
+- R102 www.ebay.com .+ |
|
529 |
+ |
|
530 |
+# Introduction to regular expressions |
|
531 |
+ |
|
532 |
+Recomended reading: |
|
533 |
+ |
|
534 |
+- http://www.regular-expressions.info/quickstart.html |
|
535 |
+ |
|
536 |
+- http://www.regular-expressions.info/tutorial.html |
|
537 |
+ |
|
538 |
+- regex(7) man-page: http://www.tin.org/bin/man.cgi?section=7\&topic=regex |
|
539 |
+ |
|
540 |
+## Special characters |
|
541 |
+ |
|
542 |
+- \[ |
|
543 |
+ |
|
544 |
+ the opening square bracket - it marks the beginning of a character class, see section[Character-classes](#Character-classes) |
|
545 |
+ |
|
546 |
+- \(\backslash\) |
|
547 |
+ |
|
548 |
+ the backslash - escapes special characters, see section [Escaping](#Escaping) |
|
549 |
+ |
|
550 |
+- ^ |
|
551 |
+ |
|
552 |
+ the caret - matches the beginning of a line (not needed in clamav regexes, this is implied) |
|
553 |
+ |
|
554 |
+- $ |
|
555 |
+ |
|
556 |
+ the dollar sign - matches the end of a line (not needed in clamav regexes, this is implied) |
|
557 |
+ |
|
558 |
+- ̇ |
|
559 |
+ |
|
560 |
+ the period or dot - matches *any* character |
|
561 |
+ |
|
562 |
+- | |
|
563 |
+ |
|
564 |
+ the vertical bar or pipe symbol - matches either of the token on its left and right side, see section [Alternation](#sub:Alternation) |
|
565 |
+ |
|
566 |
+- ? |
|
567 |
+ |
|
568 |
+ the question mark - matches optionally the left-side token, see section[Optional-matching,-and](Optional-matching,-and) |
|
569 |
+ |
|
570 |
+- \* |
|
571 |
+ |
|
572 |
+ the asterisk or star - matches 0 or more occurences of the left-side token, see section [Optional-matching,-and](Optional-matching,-and) |
|
573 |
+ |
|
574 |
+- + |
|
575 |
+ |
|
576 |
+ the plus sign - matches 1 or more occurences of the left-side token, see section [Optional-matching,-and](Optional-matching,-and) |
|
577 |
+ |
|
578 |
+- ( |
|
579 |
+ |
|
580 |
+ the opening round bracket - marks beginning of a group, see section [Groups](Groups) |
|
581 |
+ |
|
582 |
+- ) |
|
583 |
+ |
|
584 |
+ the closing round bracket - marks end of a group, see section[Groups](Groups) |
|
585 |
+ |
|
586 |
+## Character classes |
|
587 |
+ |
|
588 |
+## Escaping |
|
589 |
+ |
|
590 |
+Escaping has two purposes: |
|
591 |
+ |
|
592 |
+- it allows you to actually match the special characters themselves, for example to match the literal *+*, you would write *\(\backslash\)+* |
|
593 |
+ |
|
594 |
+- it also allows you to match non-printable characters, such as the tab (*\(\backslash\)t*), newline (*\(\backslash\)n*), .. |
|
595 |
+ |
|
596 |
+However since non-printable characters are not valid inside an url, you won’t have a reason to use them. |
|
597 |
+ |
|
598 |
+## Alternation |
|
599 |
+ |
|
600 |
+## Optional matching, and repetition |
|
601 |
+ |
|
602 |
+## Groups |
|
603 |
+ |
|
604 |
+Groups are usually used together with repetition, or alternation. For example: *(com|it)+* means: match 1 or more repetitions of *com* or *it,* that is it matches: com, it, comcom, comcomcom, comit, itit, ititcom,... you get the idea. |
|
605 |
+ |
|
606 |
+Groups can also be used to extract substring, but this is not supported by the clam engine, and not needed either in this case. |
|
607 |
+ |
|
608 |
+# How to create database files |
|
609 |
+ |
|
610 |
+## How to create and maintain the whitelist (daily.wdb) |
|
611 |
+ |
|
612 |
+If the phishing code claims that a certain mail is phishing, but its not, you have 2 choices: |
|
613 |
+ |
|
614 |
+- examine your rules daily.pdb, and fix them if necessary (see: section[How-to-create](How-to-create)) |
|
615 |
+ |
|
616 |
+- add it to the whitelist (discussed here) |
|
617 |
+ |
|
618 |
+Lets assume you are having problems because of links like this in a mail: |
|
619 |
+ |
|
620 |
+```html |
|
621 |
+ <a href=''http://69.0.241.57/bCentral/L.asp?L=XXXXXXXX''> |
|
622 |
+ http://www.bcentral.it/ |
|
623 |
+ </a> |
|
624 |
+``` |
|
625 |
+ |
|
626 |
+After investigating those sites further, you decide they are no threat, and create a line like this in daily.wdb: |
|
627 |
+ |
|
628 |
+``` |
|
629 |
+R http://www\(\backslash\).bcentral\(\backslash\).it/.+ |
|
630 |
+http://69\(\backslash\).0\(\backslash\).241\(\backslash\).57/bCentral/L\(\backslash\).asp?L=.+ |
|
631 |
+``` |
|
632 |
+ |
|
633 |
+Note: urls like the above can be used to track unique mail recipients, and thus know if somebody actually reads mails (so they can send more spam). However since this site required no authentication information, it is safe from a phishing point of view. |
|
634 |
+ |
|
635 |
+## How to create and maintain the domainlist (daily.pdb) |
|
636 |
+ |
|
637 |
+When not using –phish-scan-alldomains (production environments for example), you need to decide which urls you are going to check. |
|
638 |
+ |
|
639 |
+Although at a first glance it might seem a good idea to check everything, it would produce false positives. Particularly newsletters, ads, etc. are likely to use URLs that look like phishing attempts. |
|
640 |
+ |
|
641 |
+Lets assume that you’ve recently seen many phishing attempts claiming they come from Paypal. Thus you need to add paypal to daily.pdb: |
|
642 |
+ |
|
643 |
+``` |
|
644 |
+R .+ .+\(\backslash\).paypal\(\backslash\).com |
|
645 |
+``` |
|
646 |
+ |
|
647 |
+The above line will block (detect as phishing) mails that contain urls that claim to lead to paypal, but they don’t in fact. |
|
648 |
+ |
|
649 |
+Be carefull not to create regexes that match a too broad range of urls though. |
|
650 |
+ |
|
651 |
+## Dealing with false positives, and undetected phishing mails |
|
652 |
+ |
|
653 |
+### False positives |
|
654 |
+ |
|
655 |
+Whenever you see a false positive (mail that is detected as phishing, but its not), you need to examine *why* clamav decided that its phishing. You can do this easily by building clamav with debugging (./configure –enable-experimental –enable-debug), and then running a tool: |
|
656 |
+ |
|
657 |
+```bash |
|
658 |
+$contrib/phishing/why.py phishing.eml |
|
659 |
+``` |
|
660 |
+ |
|
661 |
+This will show the url that triggers the phish verdict, and a reason why that url is considered phishing attempt. |
|
662 |
+ |
|
663 |
+Once you know the reason, you might need to modify daily.pdb (if one of yours rules inthere are too broad), or you need to add the url to daily.wdb. If you think the algorithm is incorrect, please file a bug report on bugzilla.clamav.net, including the output of *why.py*. |
|
664 |
+ |
|
665 |
+### Undetected phish mails |
|
666 |
+ |
|
667 |
+Using why.py doesn’t help here unfortunately (it will say: clean), so all you can do is: |
|
668 |
+ |
|
669 |
+```bash |
|
670 |
+$clamscan/clamscan –phish-scan-alldomains undetected.eml |
|
671 |
+``` |
|
672 |
+ |
|
673 |
+And see if the mail is detected, if yes, then you need to add an appropriate line to daily.pdb (see section [How-to-create](How-to-create)). |
|
674 |
+ |
|
675 |
+If the mail is not detected, then try using: |
|
676 |
+ |
|
677 |
+```bash |
|
678 |
+$clamscan/clamscan –debug undetected.eml|less |
|
679 |
+``` |
|
680 |
+ |
|
681 |
+Then see what urls are being checked, see if any of them is in a whitelist, see if all urls are detected, etc. |
0 | 682 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,23 @@ |
0 |
+# Whitelist databases |
|
1 |
+ |
|
2 |
+## File whitelists |
|
3 |
+ |
|
4 |
+To whitelist a specific file use the MD5 signature format and place it inside a database file with the extension of `.fp`. To whitelist a specific file with the SHA1 or SHA256 file hash signature format, place the signature inside a database file with the extension of `.sfp`. |
|
5 |
+ |
|
6 |
+## Signature whitelists |
|
7 |
+ |
|
8 |
+To whitelist a specific signature from the database you just add the signature name into a local file with the `.ign2` extension and store it inside the database directory. |
|
9 |
+ |
|
10 |
+E.g: |
|
11 |
+ |
|
12 |
+``` |
|
13 |
+ Eicar-Test-Signature |
|
14 |
+``` |
|
15 |
+ |
|
16 |
+Additionally, you can follow the signature name with the MD5 of the entire database entry for this signature. In such a case, the signature will no longer be whitelisted when its entry in the database gets modified (eg. the signature gets updated to avoid false alerts). E.g: |
|
17 |
+ |
|
18 |
+``` |
|
19 |
+ Eicar-Test-Signature:bc356bae4c42f19a3de16e333ba3569c |
|
20 |
+``` |
|
21 |
+ |
|
22 |
+Historically, signature whitelists were added to `.ign` files. This format is still functional, though it has been replaced by the `.ign2` database. |
0 | 23 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,37 @@ |
0 |
+# Using YARA rules in ClamAV |
|
1 |
+ |
|
2 |
+ClamAV version 0.99 and above can process YARA rules. ClamAV virus database file names ending with “.yar” or “.yara” are parsed as yara rule files. The link to the YARA rule grammar documentation may be found at http://plusvic.github.io/yara/. There are currently a few limitations on using YARA rules within ClamAV: |
|
3 |
+ |
|
4 |
+- YARA modules are not yet supported by ClamAV. This includes the “import” keyword and any YARA module-specific keywords. |
|
5 |
+ |
|
6 |
+- Global rules(“global” keyword) are not supported by ClamAV. |
|
7 |
+ |
|
8 |
+- External variables(“contains” and “matches” keywords) are not supported. |
|
9 |
+ |
|
10 |
+- YARA rules pre-compiled with the *yarac* command are not supported. |
|
11 |
+ |
|
12 |
+- As in the ClamAV logical and extended signature formats, YARA strings and segments of strings separated by wild cards must represent at least two octets of data. |
|
13 |
+ |
|
14 |
+- There is a maximum of 64 strings per YARA rule. |
|
15 |
+ |
|
16 |
+- YARA rules in ClamAV must contain at least one literal, hexadecimal, or regular expression string. |
|
17 |
+ |
|
18 |
+In addition, there are a few more ClamAV processing modes that may affect the outcome of YARA rules. |
|
19 |
+ |
|
20 |
+- *File decomposition and decompression* - Since ClamAV uses file decomposition and decompression to find viruses within de-archived and uncompressed inner files, YARA rules executed by ClamAV will match against these files as well. |
|
21 |
+ |
|
22 |
+- *Normalization* - By default, ClamAV normalizes HTML, JavaScript, and ASCII text files. YARA rules in ClamAV will match against the normalized result. The effects of normalization of these file types may be captured using `clamscan --leave-temps --tempdir=mytempdir`. YARA rules may then be written using the normalized file(s) found in `mytempdir`. Alternatively, starting with ClamAV 0.100.0, `clamscan --normalize=no` will prevent normalization and only scan the raw file. To obtain similar behavior prior to 0.99.2, use `clamscan --scan-html=no`. The corresponding parameters for clamd.conf are `Normalize` and `ScanHTML`. |
|
23 |
+ |
|
24 |
+- *YARA conditions driven by string matches* - All YARA conditions are driven by string matches in ClamAV. This saves from executing every YARA rule on every file. Any YARA condition may be augmented with a string match clause which is always true, such as: |
|
25 |
+ |
|
26 |
+```yara |
|
27 |
+ rule CheckFileSize |
|
28 |
+ { |
|
29 |
+ strings: |
|
30 |
+ $abc = "abc" |
|
31 |
+ condition: |
|
32 |
+ ($abc or not $abc) and filesize < 200KB |
|
33 |
+ } |
|
34 |
+``` |
|
35 |
+ |
|
36 |
+This will ensure that the YARA condition always performs the desired action (checking the file size in this example), |