Numeric Range Comparison Without Minimum Check
|ID: 839||Date: (C)2012-05-14 (M)2012-11-08|
|Type: weakness||Status: INCOMPLETE|
|Abstraction Type: Base|
The program checks a value to ensure that it does not exceed a
maximum, but it does not verify that the value exceeds the
Extended DescriptionSome programs use signed integers or floats even when their values are
only expected to be positive or 0. An input validation check might assume
that the value is positive, and only check for the maximum value. If the
value is negative, but the code assumes that the value is positive, this can
produce an error. The error may have security consequences if the negative
value is used for memory allocation, array access, buffer access, etc.
Ultimately, the error could lead to a buffer overflow or other type of
memory corruption.The use of a negative number in a positive-only context could have
security implications for other types of resources. For example, a shopping
cart might check that the user is not requesting more than 10 items, but a
request for -3 items could cause the application to calculate a negative
price and credit the attacker's account.
Applicable PlatformsLanguage: CLanguage: OftenLanguage: C++Language: Often
|IntegrityConfidentialityAvailability ||Modify application
dataExecute unauthorized code or
commands ||An attacker could modify the structure of the message or data being
sent to the downstream component, possibly injecting commands. |
|Availability ||DoS: resource consumption
(other) ||in some contexts, a negative value could lead to resource
|ConfidentialityIntegrity ||Modify memoryRead memory ||If a negative value is used to access memory, buffers, or other
indexable structures, it could access memory outside the bounds of the
|Implementation ||Enforcement by Conversion ||If the number to be used is always expected to be positive, change the
variable type from signed to unsigned or size_t. || || |
|Implementation ||Input Validation ||If the number to be used could have a negative value based on the
specification (thus requiring a signed value), but the number should
only be positive to preserve code correctness, then include a check to
ensure that the value is positive. || || |
|CWE-839 CanPrecede CWE-124 ||Weakness ||CWE-1000 || |
Demonstrative Examples (Details)
- In the following code, the method retrieves a value from an array at
a specific array index location that is given as an input parameter to the
method (Demonstrative Example Id DX-100)
- The following code is intended to read an incoming packet from a
socket and extract one or more headers. (Demonstrative Example Id DX-21)
- The following code reads a maximum size and performs a sanity check
on that size. It then performs a strncpy, assuming it will not exceed the
boundaries of the array. While the use of "short s" is forced in this
particular example, short int's are frequently used within real-world code,
such as code that processes structured data. (Demonstrative Example Id DX-23)
- The following code shows a simple BankAccount class with deposit and
- CVE-2010-1866 : Chain: integer overflow causes a negative signed value, which later bypasses a maximum-only check, leading to heap-based buffer overflow.
- CVE-2009-1099 : Chain: 16-bit counter can be interpreted as a negative value, compared to a 32-bit maximum value, leading to buffer under-write.
- CVE-2011-0521 : Chain: kernel's lack of a check for a negative value leads to memory corruption.
- CVE-2010-3704 : Chain: parser uses atoi() but does not check for a negative value, which can happen on some platforms, leading to buffer under-write.
- CVE-2010-2530 : Chain: Negative value stored in an int bypasses a size check and causes allocation of large amounts of memory.
- CVE-2009-3080 : Chain: negative offset value to IOCTL bypasses check for maximum index, then used as an array index for buffer under-read.
- CVE-2008-6393 : chain: file transfer client performs signed comparison, leading to integer overflow and heap-based buffer overflow.
- CVE-2008-4558 : chain: negative ID in media player bypasses check for maximum index, then used as an array index for buffer under-read.
For more examples, refer to CVE relations in the bottom box.
White Box Definitions None
Black Box Definitions None
- Mark Dowd John McDonald Justin Schuh .The Art of Software Security Assessment 1st Edition. Addison Wesley. Section:'Chapter 6, "Type Conversion Vulnerabilities" Page
246.'. Published on 2006.
- Mark Dowd John McDonald Justin Schuh .The Art of Software Security Assessment 1st Edition. Addison Wesley. Section:'Chapter 6, "Comparisons", Page 265.'. Published on 2006.