Avoid performing bitwise and arithmetic operations on the same data. In particular, bitwise operations are frequently performed on arithmetic values as a form of premature optimization. Bitwise operators include the unary operator ~
and the binary operators <<
, >>
, &
, ^
, and |
. Although such operations are valid and will compile, they can reduce code readability. Declaring a variable as containing a numeric value or a bitmap makes the programmer's intentions clearer and the code more maintainable.
Bitmapped types may be defined to further separate bit collections from numeric types. Doing so may make it easier to verify that bitwise operations are performed only on variables that represent bitmaps.
typedef uint32_t bitmap32_t; bitmap32_t x = 0x000007f3; x = (x << 2) | 3; /* Shifts in two 1-bits from the right */
The typedef
name uintN_t
designates an unsigned integer type with width N
. Consequently, uint32_t
denotes an unsigned integer type with a width of exactly 32 bits. Bitmaps should be declared as unsigned. See INT13-C. Use bitwise operators only on unsigned operands.
Left- and right-shift operators are often employed to multiply or divide a number by a power of 2. However, using shift operators to represent multiplication or division is an optimization that renders the code less portable and less readable. Furthermore, most compilers routinely optimize multiplications and divisions by constant powers of 2 with bit-shift operations, and they are more familiar with the implementation details of the current platform.
Noncompliant Code Example (Left Shift)
In this noncompliant code example, both bit manipulation and arithmetic manipulation are performed on the integer x
. The result is a (prematurely) optimized statement that assigns 5x + 1
to x
for implementations where integers are represented as two's complement values.
unsigned int x = 50; x += (x << 2) + 1;
Although this is a valid manipulation, the result of the shift depends on the underlying representation of the integer type and is consequently implementation-defined. Additionally, the readability of the code is reduced.
Compliant Solution (Left Shift)
In this compliant solution, the assignment statement is modified to reflect the arithmetic nature of x
, resulting in a clearer indication of the programmer's intentions:
unsigned int x = 50; x = 5 * x + 1;
A reviewer may now recognize that the operation should also be checked for wrapping. This might not have been apparent in the original, noncompliant code example.
Noncompliant Code Example (Right Shift)
In this noncompliant code example, the programmer prematurely optimizes code by replacing a division with a right shift:
int x = -50; x >>= 2;
Although this code is likely to perform the division correctly, it is not guaranteed to. If x
has a signed type and a negative value, the operation is implementation-defined and can be implemented as either an arithmetic shift or a logical shift. In the event of a logical shift, if the integer is represented in either one's complement or two's complement form, the most significant bit (which controls the sign for both representations) will be set to 0, causing a once negative number to become a possibly very large, positive number. For more details, see INT13-C. Use bitwise operators only on unsigned operands.
For example, if the internal representation of x
is 0xFFFF FFCE
(two's complement), an arithmetic shift results in 0xFFFF FFF3
(−13 in two's complement), whereas a logical shift results in 0x3FFF FFF3
(1,073,741,811 in two's complement).
Compliant Solution (Right Shift)
In this compliant solution, the right shift is replaced by division:
int x = -50; x /= 4;
The resulting value is now more likely to be consistent with the programmer's expectations.
Exceptions
INT14-C-EX0: Routines may treat integers as bit vectors for I/O purposes. That is, they may treat an integer as a series of bits in order to write it to a file or socket. They may also read a series of bits from a file or socket and create an integer from the bits.
Risk Assessment
Performing bit manipulation and arithmetic operations on the same variable obscures the programmer's intentions and reduces readability. It also makes it more difficult for a security auditor or maintainer to determine which checks must be performed to eliminate security flaws and ensure data integrity.
Recommendation | Severity | Likelihood | Remediation Cost | Priority | Level |
---|---|---|---|---|---|
INT14-C | Medium | Unlikely | Medium | P4 | L3 |
Automated Detection
Tool | Version | Checker | Description |
---|---|---|---|
Compass/ROSE | Can detect violations of this recommendation. However, it can detect only those violations where both bitwise and arithmetic operators are used in the same expression | ||
LDRA tool suite | 9.7.1 | 585 S | Fully implemented |
Polyspace Bug Finder | R2024a | CERT C: Rec. INT14-C | Checks for bitwise and arithmetic operation on the same data (rec. fully covered) |
Related Vulnerabilities
Search for vulnerabilities resulting from the violation of this rule on the CERT website.
7 Comments
David Svoboda
I just came across the following code. It is notportable, but portability was explicitly not a requirement here, and I think it is a reasonable mixture of bitwise operators and math, and cannot be easily computed using pure math. How does this affect this rule?
David Svoboda
Another snippet of non-portable code. The
g_ntohs
macro of GLIB2 uses this code to swap low and high-order bytes of a 16-bit unsigned int:Definitely could use some security enhancements (to prevent overflow), but the point is: our rec here needs some way of indicating legit exceptions.
Masaki Kubo
In the automated detection section, Fortify: "... but cannot distinguish between operations of positive and negative numbers." doesn't make sense to me.
Robert Seacord
Most of the rest of that sentence didn't make sense to me either so I trimmed it down quite a lot.
David Svoboda
Another exception to this rec: The code examples in INT17-C. Define integer constants in an implementation-independent manner combine bitwise operators and addition to change the sign on a number using twos-complement arithmetic.
David Svoboda
Having found counterexamples over the years, I think we can limit them to two exceptions:
David Svoboda
INT36-C. Do not perform bitwise and arithmetic operations on the same variable renders this rule obsolete.