CWE-94: Improper Control of Generation of Code ('Code Injection')

Description

The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment.

Extended Description

N/A

ThreatScore

Threat Mapped score: 0.0

Industry: Finiancial

Threat priority: Unclassified

Observed Examples (CVEs)

CVE: CVE-2023-29374

Math component in an LLM framework translates user input into a Python expression that is input into the Python exec() method, allowing code execution - one variant of a "prompt injection" attack.
CVE: CVE-2024-5565

Python-based library uses an LLM prompt containing user input to dynamically generate code that is then fed as input into the Python exec() method, allowing code execution - one variant of a "prompt injection" attack.
CVE: CVE-2024-4181

Framework for LLM applications allows eval injection via a crafted response from a hosting provider.
CVE: CVE-2022-2054

Python compiler uses eval() to execute malicious strings as Python code.
CVE: CVE-2021-22204 — KEV

Chain: regex in EXIF processor code does not correctly determine where a string ends (CWE-625), enabling eval injection (CWE-95), as exploited in the wild per CISA KEV.
CVE: CVE-2020-8218 — KEV

"Code injection" in VPN product, as exploited in the wild per CISA KEV.
CVE: CVE-2008-5071

Eval injection in PHP program.
CVE: CVE-2002-1750

Eval injection in Perl program.
CVE: CVE-2008-5305

Eval injection in Perl program using an ID that should only contain hyphens and numbers.
CVE: CVE-2002-1752

Direct code injection into Perl eval function.
CVE: CVE-2002-1753

Eval injection in Perl program.
CVE: CVE-2005-1527

Direct code injection into Perl eval function.
CVE: CVE-2005-2837

Direct code injection into Perl eval function.
CVE: CVE-2005-1921

MFV. code injection into PHP eval statement using nested constructs that should not be nested.
CVE: CVE-2005-2498

MFV. code injection into PHP eval statement using nested constructs that should not be nested.
CVE: CVE-2005-3302

Code injection into Python eval statement from a field in a formatted file.
CVE: CVE-2007-1253

Eval injection in Python program.
CVE: CVE-2001-1471

chain: Resultant eval injection. An invalid value prevents initialization of variables, which can be modified by attacker and later injected into PHP eval statement.
CVE: CVE-2002-0495

Perl code directly injected into CGI library file from parameters to another CGI program.
CVE: CVE-2005-1876

Direct PHP code injection into supporting template file.
CVE: CVE-2005-1894

Direct code injection into PHP script that can be accessed by attacker.
CVE: CVE-2003-0395

PHP code from User-Agent HTTP header directly inserted into log file implemented as PHP script.

Related Attack Patterns (CAPEC)

Attack TTPs

T1027.009 — Embedded Payloads (defense-evasion)
T1564.009 — Resource Forking (defense-evasion)
T1027.006 — HTML Smuggling (defense-evasion)

Malware

APTs (Intrusion Sets)

Modes of Introduction

Phase	Note
Implementation	REALIZATION: This weakness is caused during implementation of an architectural security tactic.

Common Consequences

Impact: Bypass Protection Mechanism — Notes: In some cases, injectable code controls authentication; this may lead to a remote vulnerability.
Impact: Gain Privileges or Assume Identity — Notes: Injected code can access resources that the attacker is directly prevented from accessing.
Impact: Execute Unauthorized Code or Commands — Notes: When a product allows a user's input to contain code syntax, it might be possible for an attacker to craft the code in such a way that it will alter the intended control flow of the product. As a result, code injection can often result in the execution of arbitrary code. Code injection attacks can also lead to loss of data integrity in nearly all cases, since the control-plane data injected is always incidental to data recall or writing.
Impact: Hide Activities — Notes: Often the actions performed by injected control code are unlogged.

Potential Mitigations

Architecture and Design: Refactor your program so that you do not have to dynamically generate code. (N/A)
Architecture and Design: Run your code in a "jail" or similar sandbox environment that enforces strict boundaries between the process and the operating system. This may effectively restrict which code can be executed by your product. Examples include the Unix chroot jail and AppArmor. In general, managed code may provide some protection. This may not be a feasible solution, and it only limits the impact to the operating system; the rest of your application may still be subject to compromise. Be careful to avoid CWE-243 and other weaknesses related to jails. (N/A)
Implementation: Assume all input is malicious. Use an "accept known good" input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does. When performing input validation, consider all potentially relevant properties, including length, type of input, the full range of acceptable values, missing or extra inputs, syntax, consistency across related fields, and conformance to business rules. As an example of business rule logic, "boat" may be syntactically valid because it only contains alphanumeric characters, but it is not valid if the input is only expected to contain colors such as "red" or "blue." Do not rely exclusively on looking for malicious or malformed inputs. This is likely to miss at least one undesirable input, especially if the code's environment changes. This can give attackers enough room to bypass the intended validation. However, denylists can be useful for detecting potential attacks or determining which inputs are so malformed that they should be rejected outright. To reduce the likelihood of code injection, use stringent allowlists that limit which constructs are allowed. If you are dynamically constructing code that invokes a function, then verifying that the input is alphanumeric might be insufficient. An attacker might still be able to reference a dangerous function that you did not intend to allow, such as system(), exec(), or exit(). (N/A)
Testing: Use automated static analysis tools that target this type of weakness. Many modern techniques use data flow analysis to minimize the number of false positives. This is not a perfect solution, since 100% accuracy and coverage are not feasible. (N/A)
Testing: Use dynamic tools and techniques that interact with the product using large test suites with many diverse inputs, such as fuzz testing (fuzzing), robustness testing, and fault injection. The product's operation may slow down, but it should not become unstable, crash, or generate incorrect results. (N/A)
Operation: Run the code in an environment that performs automatic taint propagation and prevents any command execution that uses tainted variables, such as Perl's "-T" switch. This will force the program to perform validation steps that remove the taint, although you must be careful to correctly validate your inputs so that you do not accidentally mark dangerous inputs as untainted (see CWE-183 and CWE-184). (N/A)
Operation: Run the code in an environment that performs automatic taint propagation and prevents any command execution that uses tainted variables, such as Perl's "-T" switch. This will force the program to perform validation steps that remove the taint, although you must be careful to correctly validate your inputs so that you do not accidentally mark dangerous inputs as untainted (see CWE-183 and CWE-184). (N/A)
Implementation: For Python programs, it is frequently encouraged to use the ast.literal_eval() function instead of eval, since it is intentionally designed to avoid executing code. However, an adversary could still cause excessive memory or stack consumption via deeply nested structures [REF-1372], so the python documentation discourages use of ast.literal_eval() on untrusted data [REF-1373]. (Discouraged Common Practice)

Applicable Platforms

None (Interpreted, Sometimes)

Demonstrative Examples

Intro: This example attempts to write user messages to a message file and allow users to view them.

Body: While the programmer intends for the MessageFile to only include data, an attacker can provide a message such as:

$MessageFile = "messages.out"; if ($_GET["action"] == "NewMessage") { $name = $_GET["name"]; $message = $_GET["message"]; $handle = fopen($MessageFile, "a+"); fwrite($handle, "<b>$name</b> says '$message'<hr>\n"); fclose($handle); echo "Message Saved!<p>\n"; } else if ($_GET["action"] == "ViewMessages") { include($MessageFile); }

Intro: edit-config.pl: This CGI script is used to modify settings in a configuration file.

Body: The script intends to take the 'action' parameter and invoke one of a variety of functions based on the value of that parameter - config_file_add_key(), config_file_set_key(), or config_file_delete_key(). It could set up a conditional to invoke each function separately, but eval() is a powerful way of doing the same thing in fewer lines of code, especially when a large number of functions or variables are involved. Unfortunately, in this case, the attacker can provide other values in the action parameter, such as:

use CGI qw(:standard); sub config_file_add_key { my ($fname, $key, $arg) = @_; # code to add a field/key to a file goes here } sub config_file_set_key { my ($fname, $key, $arg) = @_; # code to set key to a particular file goes here } sub config_file_delete_key { my ($fname, $key, $arg) = @_; # code to delete key from a particular file goes here } sub handleConfigAction { my ($fname, $action) = @_; my $key = param('key'); my $val = param('val'); # this is super-efficient code, especially if you have to invoke # any one of dozens of different functions! my $code = "config_file_$action_key(\$fname, \$key, \$val);"; eval($code); } $configfile = "/home/cwe/config.txt"; print header; if (defined(param('action'))) { handleConfigAction($configfile, param('action')); } else { print "No action specified!\n"; }

Intro: This simple script asks a user to supply a list of numbers as input and adds them together.

Body: The eval() function can take the user-supplied list and convert it into a Python list object, therefore allowing the programmer to use list comprehension methods to work with the data. However, if code is supplied to the eval() function, it will execute that code. For example, a malicious user could supply the following string:

def main(): sum = 0 numbers = eval(input("Enter a space-separated list of numbers: ")) for num in numbers: sum = sum + num print(f"Sum of {numbers} = {sum}") main()

Notes

Theoretical: Injection problems encompass a wide variety of issues -- all mitigated in very different ways. For this reason, the most effective way to discuss these weaknesses is to note the distinct features that classify them as injection weaknesses. The most important issue to note is that all injection problems share one thing in common -- i.e., they allow for the injection of control plane data into the user-controlled data plane. This means that the execution of the process may be altered by sending code in through legitimate data channels, using no other mechanism. While buffer overflows, and many other flaws, involve the use of some further issue to gain execution, injection problems need only for the data to be parsed. The most classic instantiations of this category of weakness are SQL injection and format string vulnerabilities.

← Back to CWE list