When creating your rules for YARA keep in mind the following guidelines in order to get the best performance from them. This guide is based on ideas and recommendations by Victor M. Alvarez and WXS.
- Revision 1.4, October 2020, applies to all YARA versions higher than 3.7
To get a better grip on what and where YARA performance can be optimized, it's useful to understand the scanning process. It's basically separated into 4 steps which will be explained very simplified using this examples rule:
import "math"
rule example_php_webshell_rule
{
meta:
description = "Just an example php webshell rule"
date = "2021/02/16"
strings:
$php_tag = "<?php"
$input1 = "GET"
$input2 = "POST"
$payload = /assert[\t ]{0,100}\(/
condition:
filesize < 20KB and
$php_tag and
$payload and
any of ( $input* ) and
math.entropy(500, filesize-500) >= 5
}
This step happens before the actual scan. YARA will look for so called atoms in the search strings to feed the Aho-Corasick automaton. The details are explained in the chapter atom but for now it's enough to know, that they're maximum 4 bytes longs and YARA picks them quite cleverly to avoid too many matches. In our example YARA might pick the following 4 atoms:
<?phGETPOSTsser(out ofassert)
Here the scan has started. Steps 2.-4. will be executed on all files. YARA will look in each file for the 4 atoms defined above. Any matches are handed over to the bytecode engine.
If there's e.g. a match on sser, YARA will check if it was prefixed by an a and continues with a t. If that is true, it will follow on with the regex [\t ]{0,100}\(. With this clever approach YARA avoids going with a slow regex engine over the complete files and just picks certain parts to look closer.
After all pattern matching is done, the conditions are checked.
YARA has another optimization mechanism to only do the CPU intense math.entropy check from our example rule, if the 4 conditions before it are satisfied. Explained in more details in the chapter Conditions and Short-Circuit Evaluation
If the conditions are satisfied, a match is reported. The scan continues with the next file in step 2.
YARA extracts from the strings short substrings up to 4 bytes long that are called "atoms". Those atoms can be extracted from any place within the string, and YARA searches for those atoms while scanning the file, if it finds one of the atoms then it verifies that the string actually matches.
For example, consider this strings:
/abc.*cde/
=> possible atoms are abc and cde, either one or the other can be used
/(one|two)three/
=> possible atoms are one, two and three, we can search for three alone, or for both one and two
YARA does its best effort to select the best atoms from each string, for example:
{ 00 00 00 00 [1-4] 01 02 03 04 }
=> here YARA uses the atom 01 02 03 04, because 00 00 00 00 is too common
{ 01 02 [1-4] 01 02 03 04 }
=> 01 02 03 04 is preferred over 01 02 because it's longer
So, the important point is that strings should contain good atoms. These are bad strings because they contain either too short or too common atoms:
{00 00 00 00 [1-2] FF FF [1-2] 00 00 00 00}
{AB [1-2] 03 21 [1-2] 01 02}
/a.*b/
/a(c|d)/
The worst strings are those that don't contain any atoms at all, like:
/\w.*\d/
/[0-9]+\n/
This regular expression don't contain any fixed substring that can be used as atom, so it must be evaluated at every offset of the file to see if it matches there.
Another good import recommendation is to avoid for loops with too many iterations, specially of the statement within the loop is too complex, for example:
strings:
$a = {00 00}
condition:
for all i in (1..#a) : (@a[i] < 10000)
This rule has two problems. The first is that the string $a is too common, the second one is that because $a is too common #a can be too high and can be evaluated thousands of times.
This other condition is also inefficient because the number of iterations depends on filesize, which can be also very high:
for all i in (1..filesize) : ($a at i)
Avoid using the "magic" module which is not available on the Windows platform. Using the "magic" module slows down scanning but provides exact matches.
Custom GIF magic header definition:
rule gif_1 {
condition:
(uint32be(0) == 0x47494638 and uint16be(4) == 0x3961) or
(uint32be(0) == 0x47494638 and uint16be(4) == 0x3761)
}
Using the "magic" module:
import "magic"
rule gif_2 {
condition:
magic.mime_type() == "image/gif"
}
Avoid defining too short strings. Any string with less than 4 bytes will probably appear in a lot of files OR as uniform content in an XORed file.
Some strings are long enough but shouldn't be used due to a different reason - uniformity. These are some examples for strings that shouldn't be used as they could cause too many matches in files.
$s1 = "22222222222222222222222222222222222222222222222222222222222222" ascii
$s2 = "\x00\x20\x00\x20\x00\x20\x00\x20\x00\x20\x00\x20\x00\x20" ascii // wide formatted spaces
Error message would look like:
error scanning yara-killer.dat: string "$mz" in rule "shitty_mz" caused too many matches
Try to describe string definitions as narrow as possible. Avoid the "nocase" attribute if possible, because many atoms will be generated and searched for (higher memory usage, more iterations). Remember, in the absence of modifiers "ascii" is assumed by default. The possible combinations are:
LOW - only one atom is generated
$s1 = "cmd.exe" // (ascii only)
$s2 = "cmd.exe" ascii // (ascii only, same as $s1)
$s3 = "cmd.exe" wide // (UTF-16 only)
$s4 = "cmd.exe" ascii wide // (both ascii and UTF-16) two atoms will be generated
$s5 = { 63 6d 64 2e 65 78 65 } // ascii char code in hex
HIGH - All combinations of upper and lowercase letters for the 4 bytes chosen by YARA will be generated as atoms
$s5 = "cmd.exe" nocase (all different cases, e.g. "Cmd.", "cMd.", "cmD." ..)
If you want to match scripting commands, check if the language is case insensitive at all (e.g. php, Windows batch) before using nocase. If you just need different casing for just one or two letters, you're better off with a regex, e.g.
$ = /[Pp]assword/
Use expressions only when necessary. Regular expression evaluation is inherently slower than plain string matching and consumes a significant amount of memory. Don't use them if hex strings with jumps and wild-cards can solve the problem.
If you have to use regular expressions avoid greedy .* and even reluctant quantifiers .*?. Instead use exact numbers like .{1,30} or even .{1,3000}.
If you want to make sure, that e.g. exec is followed by /bin/sh, you can use the offsets supplied by the @ symbol. This would be the slow regex version:
$ = /exec.*\/bin\/sh/
This is the faster offset way:
strings:
$exec = "exec"
$sh = "/bin/sh"
conditions:
$exec and $sh and
@exec < @sh
Also try to include long sequences of strings that could serve as anchors in the matching progress. Again, the longer the better.
BAD
$s1 = /http:\/\/[.]*\.hta/
BETTER
$s1 = /http:\/\/[a-z0-9\.\/]{3,70}\.hta/
BEST
$s1 = /mshta\.exe http:\/\/[a-z0-9\.\/]{3,70}\.hta/
Try to write condition statements in which the elements that are most likely to be "False" are placed first. The condition is evaluated from left to right. The sooner the engine identifies that a rule is not satisfied the sooner it can skip the current rule and evaluate the next one. The speed improvement caused by this way to order the condition statements depends on the difference in necessary CPU cycles to process each of the statements. If all statements are more or less equally expensive, reordering the statements causes no noticeable improvement. If one of the statements can be processed very fast it is recommended to place it first in order to skip the expensive statement evaluation in cases in which the first statement is FALSE.
Changing the order in the following statement does not cause a significant improvement:
$string1 and $string2 and uint16(0) == 0x5A4D
However, if the execution time of the statements is very different, reordering in order to trigger the short-circuit will improve the scan speed significantly:
SLOW
EXPENSIVE and CHEAP
math.entropy(0, filesize) > 7.0 and uint16(0) == 0x5A4D
FAST
CHEAP and EXPENSIVE
uint16(0) == 0x5A4D and math.entropy(0, filesize) > 7.0
Short-circuit evaluation was introduced to help optimizing expensive sentences, particularly "for" sentences. Some people were using conditions like the one in the following example:
strings:
$mz = "MZ"
...
condition:
$mz at 0 and for all i in (1..filesize) : ( whatever )
Because filesize can be a very big number, "whatever" can be executed a lot of times, slowing down the execution. Now, with short-circuit evaluation, the "for" sentence will be executed only if the first part of the condition is met, so, this rule will be slow only for MZ files. An additional improvement could be:
$mz at 0 and filesize < 100K and for all i in (1..filesize) : ( whatever )
This way a higher bound to the number of iterations is set.
Sadly this does not work with regular expressions because they're all initially fed into the string matching engine. The following example will slow down the search for any file and not just for those with filesize smaller than 200 bytes:
strings:
$expensive_regex = /\$[a-z0-9_]+\(/ nocase
conditions:
filesize < 200 and
$expensive_regex
This "short-circuit" evaluation is applied since YARA version 3.4.
Any data in the metadata section is read into the RAM by YARA. (You can easily test this by inserting 100,000 hashes into a rule and check the RAM usage of the YARA scan before and after.) Of course you don't want to permanently remove the metadata from the rules but if you're short on RAM, you could remove some unneeded parts of it in your workflow directly prior to scanning.