Secure Coding in C

and C++
Integer Security

Lecture 7
Acknowledgement: These slides are based on author Seacord’s original presentation
Integer Security
 Integers represent a growing and underestimated
source of vulnerabilities in C and C++ programs.
 Integer range checking has not been systematically
applied in the development of most C and C++
software.
 security flaws involving integers exist
 a portion of these are likely to be vulnerabilities

 A software vulnerability may result when a program

evaluates an integer to an unexpected value.
Integer Representation
 Signed-magnitude

 One’s complement
 Two’s complement

 These integer representations vary in how

they represent negative numbers

Signed-magnitude Representation
 Uses the high-order bit to indicate the sign
0 for positive
1 for negative
remaining low-order bits indicate the magnitude of the
value
0 0 1 0 1 0 0 1 1 0 1 0 1 0 0 1
32 + 8 + 1 32 + 8 + 1

+ 41 - 41
 Signed magnitude representation of +41 and -
41
One’s Complement
 One’s complement replaced signed
magnitude because the circuitry was too
complicated.
 Negative numbers are represented in one’s

complement form by complementing each bit

0 0 1 0 1 0 0 1 each 1 is
even the
replaced
sign bit is
1 1 0 1 0 1 1 0 with a 0
reversed

each 0 is
replaced
with a 1
Two’s Complement
 The two’s complement form of a negative integer is created by
adding one to the one’s complement representation.
0 0 1 0 1 0 0 1 0 0 1 0 1 0 0 1

1 1 0 1 0 1 1 0+ 1 = 1 1 0 1 0 1 1 1
 Two’s complement representation has a single (positive) value
for zero.
 The sign is represented by the most significant bit.
 The notation for positive integers is identical to their signed-
magnitude representations.
Signed and Unsigned Types
 Integers in C and C++ are either signed or unsigned.
 Signed integers
 represent positive and negative values.
 In two’s complement arithmetic, a signed integer ranges
from -2n-1 through 2n-1-1.
 Unsigned integers
 range from zero to a maximum that depends on the size of
the type
 This maximum value can be calculated as
2n-1, where n is the number of bits used to represent the
unsigned type.
Representation

4-bit
two’s complement
representation

Signed Integer Unsigned Integer

Standard Integer Types
 Standardintegers include the following types,
in non-decreasing length order
signed char
short int

int

long int

long long int

Platform-Specific Integer
Types
 Vendors often define platform-specific integer types.
 The Microsoft Windows API defines a large number of
integer types
 __int8, __int16, __int32, __int64
 ATOM
 BOOLEAN, BOOL
 BYTE
 CHAR
 DWORD, DWORDLONG, DWORD32, DWORD64
 WORD
 INT, INT32, INT64
 LONG, LONGLONG, LONG32, LONG64
 Etc.
Integer Ranges
 Minimum and maximum values for an integer
type depend on
 the type’s representation
 signedness
 number of allocated bits
 TheC99 standard sets minimum
requirements for these ranges.
Example Integer Ranges
signed char

-128 0 127
unsigned char

short 0 255

- 32768 0 32767
unsigned short

0 65535
Integer Conversions
 Type conversions
 occur explicitly in C and C++ as the result of a cast or
 implicitly as required by an operation.
 Conversions can lead to lost or misinterpreted data.
 Implicit conversions are a consequence of the C language
ability to perform operations on mixed types.
 C99 rules define how C compilers handle
conversions
 integer promotions
 integer conversion rank
 usual arithmetic conversions
Integer Promotions
 Integer types smaller than int are promoted
when an operation is performed on them.
 If all values of the original type can be

represented as an int
 thevalue of the smaller type is converted to int
 otherwise, it is converted to unsigned int.

 Integerpromotions are applied as part of the

usual arithmetic conversions
Integer Promotion Example
 Integerpromotions require the promotion of
each variable (c1 and c2) to int size
char c1, c2;
c1 = c1 + c2;
 The two ints are added and the sum truncated to fit
into the char type.
 Integer promotions avoid arithmetic errors from the

overflow of intermediate values.

Implicit Conversions
The sum of c1 and c2 exceeds the
maximum size of signed char

1. char cresult, c1, c2, c3;

2. c1 = 100;
However, c1, c1, and c3 are each
3. c2 = 90; converted to integers and the overall
expression is successfully evaluated.
4. c3 = -120;
5. cresult = c1 + c2 + c3;
The sum is truncated and The value of c1 is added
stored in cresult without a to the value of c2.
loss of data
Integer Conversion Rank &
Rules
 Every integer type has an integer conversion
rank that determines how conversions are
performed.
 No two signed integer types have the same rank,
even if they have the same representation.
 The rank of a signed integer type is > the rank of
any signed integer type with less precision.
 rank of [long long int > long int> int > short
int > signed char].
 The rank of any unsigned integer type is equal to
the rank of the corresponding signed integer type.
Unsigned Integer Conversions
1

 Conversions of smaller unsigned integer types to

larger unsigned integer types is
 always safe
 typically accomplished by zero-extending the value

 When a larger unsigned integer is converted to a

smaller unsigned integer type the
 largervalue is truncated
 low-order bits are preserved
Unsigned Integer Conversions
2

 When unsigned integer types are converted

to the corresponding signed integer type
 the bit pattern is preserved so no data is lost
 the high-order bit becomes the sign bit

 If the sign bit is set, both the sign and magnitude of the

value changes.
 From To Method
unsigned
char char Preserve bit pattern; high-order bit becomes sign bit
char short Zero-extend
char long Zero-extend
char unsigned Zero-extend
short
char unsigned long Zero-extend
short char Preserve low-order byte
short short Preserve bit pattern; high-order bit becomes sign bit
short long Zero-extend
short unsigned char Preserve low-order byte
long char Preserve low-order byte
long short Preserve low-order word
long long Preserve bit pattern; high-order bit becomes sign bit
long unsigned char Preserve low-order byte
long unsigned Preserve low-order word
short
Key: Lost data Misinterpreted data
Signed Integer Conversions 1
 When a signed integer is converted to an
unsigned integer of equal or greater size and
the value of the signed integer is not negative
 the value is unchanged
 the signed integer is sign-extended

A signed integer is converted to a shorter

signed integer by truncating the high-order
bits.
Signed Integer Conversions 2
 When signed integers are converted to
unsigned integers
 bit pattern is preserved—no lost data
 high-order bit loses its function as a sign bit

 If the value of the signed integer is not negative, the

value is unchanged.
 If the value is negative, the resulting unsigned value is

evaluated as a large, signed integer.

 From To Method
char short Sign-extend
char long Sign-extend
char unsigned char Preserve pattern; high-order bit loses function as sign bit
char unsigned short Sign-extend to short; convert short to unsigned short
char unsigned long Sign-extend to long; convert long to unsigned long
short char Preserve low-order byte
short long Sign-extend
short unsigned char Preserve low-order byte
short unsigned short Preserve bit pattern; high-order bit loses function as sign
bit
short unsigned long Sign-extend to long; convert long to unsigned long
long char Preserve low-order byte
long short Preserve low-order word
long unsigned char Preserve low-order byte
long unsigned short Preserve low-order word
long unsigned long Preserve pattern; high-order bit loses function as sign bit

Key: Lost data Misinterpreted data

Signed Integer Conversion
Example
 1. unsigned int l = ULONG_MAX;
The value of c is
 2. char c = -1; compared to the
 3. if (c == l) { value of l.
 4. printf("-1 = 4,294,967,295?\n");
 5. }

Because of integer promotions, c is

converted to an unsigned integer with a
value of 0xFFFFFFFF or 4,294,967,295
Signed/Unsigned Characters
 The type char can be signed or unsigned.
 When a signed char with its high bit set is

saved in an integer, the result is a negative

number.
 Use unsigned char for buffers, pointers,

and casts when dealing with character data

that may have values greater than 127
(0x7f).
Usual Arithmetic Conversions
 If both operands have the same type no conversion is needed.
 If both operands are of the same integer type (signed or unsigned),
the operand with the type of lesser integer conversion rank is
converted to the type of the operand with greater rank.
 If the operand that has unsigned integer type has rank >= to the rank
of the type of the other operand, the operand with signed integer type
is converted to the type of the operand with unsigned integer type.
 If the type of the operand with signed integer type can represent all of
the values of the type of the operand with unsigned integer type, the
operand with unsigned integer type is converted to the type of the
operand with signed integer type.
 Otherwise, both operands are converted to the unsigned integer type
corresponding to the type of the operand with signed integer type.
Integer Error Conditions
 Integeroperations can resolve to unexpected
values as a result of an
 overflow
 sign error
 truncation
Overflow
 An integer overflow occurs when an integer is
increased beyond its maximum value or
decreased beyond its minimum value.
 Overflows can be signed or unsigned

A signed overflow An unsigned overflow

occurs when a value is occurs when the underlying
carried over to the sign representation can no longer
bit represent a value
Overflow Examples 1
 1. int i;
 2. unsigned int j;

 3. i = INT_MAX; // 2,147,483,647
 4. i++;
 5. printf("i = %d\n", i); i=-2,147,483,648

 6. j = UINT_MAX; // 4,294,967,295;
 7. j++;
 8. printf("j = %u\n", j);
j = 0
Overflow Examples 2
 9. i = INT_MIN; // -2,147,483,648;
 10. i--; i=2,147,483,647
 11. printf("i = %d\n", i);

 12. j = 0;
 13. j--;
 14. printf("j = %u\n", j);
j = 4,294,967,295
Truncation Errors
 Truncation errors occur when
 an integer is converted to a smaller integer type
and
 the value of the original integer is outside the
range of the smaller type
 Low-orderbits of the original value are
preserved and the high-order bits are lost.
Truncation Error Example
 1. char cresult, c1, c2, c3;
 2. c1 = 100; Adding c1 and c2 exceeds the max
size of signed char (+127)
 3. c2 = 90;

 4. cresult = c1 + c2;

Truncation occurs when the Integers smaller than int are

value is assigned to a type promoted to int or
that is too small to represent unsigned int before being
the resulting value operated on
Sign Errors:
Converting to Signed Integer
 Converting an unsigned integer to a signed
integer of
 Equal size - preserve bit pattern; high-order bit
becomes sign bit
 Greater size - the value is zero-extended then
converted
 Lesser size - preserve low-order bits
 If the high-order bit of the unsigned integer is
 Not set - the value is unchanged
 Set - results in a negative value
Converting to Unsigned
Integer
 Converting a signed integer to an unsigned
integer of
 Equal size - bit pattern of the original integer is
preserved
 Greater size - the value is sign-extended then converted

 Lesser size - preserve low-order bits

 If the value of the signed integer is

 Notnegative - the value is unchanged
 Negative - a (typically large) positive value
Sign Error Example
 1. int i = -3;
 2. unsigned short u; Implicit conversion
to smaller unsigned
integer
 3. u = i;
 4. printf("u = %hu\n", u);

There are sufficient bits to represent the value so

no truncation occurs. The two’s complement
representation is interpreted as a large signed
value, however, so u = 65533
Integer Operations
 Integer
operations can result in errors and
unexpected value.
 Unexpected integer values can cause
 unexpected program behavior
 security vulnerabilities
 Most integer operations can result in
exceptional conditions.
Integer Addition
 Addition can be used to add two arithmetic
operands or a pointer and an integer.
 If both operands are of arithmetic type, the

usual arithmetic conversions are performed

on them.
 Integer addition can result in an overflow if

the sum cannot be represented in the number

allocated bits
Add Instruction
 IA-32 instruction set includes an add instruction that
takes the form
 add destination, source
 Adds the 1st (destination) op to the 2nd (source) op
 Stores the result in the destination operand
 Destination operand can be a register or memory location
 Source operand can be an immediate, register, or memory
location
 Signed and unsigned overflow conditions are detected
and reported.
Add Instruction Example
 The instruction:
 add ax, bx
 adds the 16-bit bx register to the 16-bit ax register
 leaves the sum in the ax register
 The add instruction sets flags in the flags register
 overflow flag indicates signed arithmetic overflow
 carry flag indicates unsigned arithmetic overflow
Layout of the Flags Register
15 0

Overflow
Direction
Interrupt

Sign
Zero

Auxiliary Carry

Parity

Carry
Interpreting Flags
 There are no distinctions between the
addition of signed and unsigned integers at
the machine level.
 Overflow and carry flags must be interpreted

in context
Adding signed and unsigned
int
 Bothsigned int and unsigned int
values are added as follows:
 ui1 + ui2
 7. mov eax, dword ptr [ui1]
 8. add eax, dword ptr [ui2]
Adding signed long long int
The add instruction adds the
low-order 32 bits
 sll1 + sll2
 9. mov eax, dword ptr [sll1]
 10. add eax, dword ptr [sll2]
 11. mov ecx, dword ptr [ebp-98h]
 12. adc ecx, dword ptr [ebp-0A8h]

The adc instruction adds the high-order

32 bits and the value of the carry bit
Unsigned Overflow Detection
 The carry flag denotes an unsigned arithmetic
overflow
 Unsigned overflows can be detected using the
 jc instruction (jump if carry)
 jnc instruction (jump if not carry)
 Conditional jump instructions are placed after the
 add instruction in the 32-bit case
 adc instruction in the 64-bit case
Signed Overflow Detection
 The overflow flag denotes a signed arithmetic
overflow
 Signed overflows can be detected using the
 jo instruction (jump if overflow)
 jno instruction (jump if not overflow)
 Conditional jump instructions are placed after
the
 add instruction in the 32-bit case
 adc instruction in the 64-bit case
Integer Subtraction
 The IA-32 instruction set includes
 sub (subtract)
 sbb (subtract with borrow).
 The sub and sbb instructions set the overflow and
carry flags to indicate an overflow in the signed or
unsigned result.
sub Instruction
 Subtracts the 2nd (source) operand from the 1st
(destination) operand
 Stores the result in the destination operand
 The destination operand can be a
 register
 memory location
 The source operand can be a(n)
 immediate
 register
 memory location
sbb Instruction
 The sbb instruction is executed as part of a multi-
byte or multi-word subtraction.
 The sbb instruction adds the 2nd (source) operand
and the carry flag and subtracts the result from the
1st (destination) operand
 The result of the subtraction is stored in the
destination operand.
 The carry flag represents a borrow from a previous
subtraction.
signed long long int Sub
The sub instruction subtracts
the low-order 32 bits
 sll1 - sll2
 1. mov eax, dword ptr [sll1]
 2. sub eax, dword ptr [sll2]
 3. mov ecx, dword ptr [ebp-0E0h]
 4. sbb ecx, dword ptr [ebp-0F0h]

The sbb instruction subtracts the high-order 32 bits

NOTE: Assembly Code Generated by Visual C++ for Windows 2000
Integer Multiplication
 Multiplication is prone to overflow errors
because relatively small operands can
overflow
 One solution is to allocate storage for the

product that is twice the size of the larger of

the two operands.
Signed/Unsigned Examples
 The max value for an unsigned integer is 2n-1
 2n-1 x 2n-1 = 22n – 2n+1 + 1 < 22n
 The minimum value for a signed integer is -
2n-1
 -2n-1 x -2n-1 = 22n-2 2 < 22n
Multiplication Instructions
 The IA-32 instruction set includes a
 mul (unsigned multiply) instruction
 imul (signed multiply) instruction
 The mul instruction
 performs an unsigned multiplication of the 1 st
(destination) operand and the 2nd (source)
operand
 stores the result in the destination operand.
Unsigned Multiplication
Product of 8-bit operands
 1. if (OperandSize == 8) { are stored in 16-bit

destination registers
2. AX = AL * SRC;
 3. else {
 4. if (OperandSize == 16) {
 5. DX:AX = AX * SRC; Product of 16-bit operands
 are stored in 32-bit
6. } destination registers
 7. else { // OperandSize == 32
 8. EDX:EAX = EAX * SRC;
 9. }
 10. }
Product of 32-bit operands are stored in 64-bit
destination registers
Signed/Unsigned int
Multiplication
 si_product = si1 * si2;
 ui_product = ui1 * ui2;
 9. mov eax, dword ptr [ui1]
 10. imul eax, dword ptr [ui2]
 11. mov dword ptr [ui_product], eax
Upcasting
 Cast both operands to the next larger size
and then multiply.
 For unsigned integers
 check high-order bits in the next larger integer
 if any are set, throw an error.
 For signed integers all zeros or all ones in the
high-order bits and the sign bit on the low-
order bit indicate no overflow.
Upcast Example
 void* AllocBlocks(size_t cBlocks) {
 // allocating no blocks is an error
if (cBlocks == 0) return NULL;
 // Allocate enough memory
// Upcast the result to a 64-bit integer
// and check against 32-bit UINT_MAX
// to makes sure there's no overflow
 unsigned long long alloc = cBlocks * 16;
return (alloc < UINT_MAX)
 ? malloc(cBlocks * 16)
: NULL;
 Multiplication results in a 32-bit value. The result is
} assigned to a unsigned long long but the
calculation may have already overflowed.
Standard Compliance
 To be compliant with C99, multiplying two
32-bit numbers in this context must yield a
32-bit result.
 The language was not modified because the

result would be burdensome on architectures

that do not have widening multiply
instructions.
 The correct result could be achieved by

casting one of the operands.

Corrected Upcast Example
 void* AllocBlocks(size_t cBlocks) {
 // allocating no blocks is an error
if (cBlocks == 0) return NULL;
 // Allocate enough memory
// Upcast the result to a 64-bit integer
// and check against 32-bit UINT_MAX
// to makes sure there's no overflow
 unsigned long long alloc =
(unsigned long long)cBlocks*16;
return (alloc < UINT_MAX)
 ? malloc(cBlocks * 16)
: NULL;
 }
Integer Division
 An integer overflow condition occurs when the
minimum integer value for 32-bit or 64-bit
integers are divided by -1.
 In the 32-bit case, –2,147,483,648/-1 should be
equal to 2,147,483,648

- 2,147,483,648 /-1 = - 2,147,483,648

 Because 2,147,483,648 cannot be represented as
a signed 32-bit integer the resulting value is
incorrect
Error Detection
 The IA-32 instruction set includes the div
and idiv instructions
 The div instruction
 divides the (unsigned) integer value in the ax,
dx:ax, or edx:eax registers (dividend) by the
source operand (divisor)
 stores the result in the ax (ah:al), dx:ax, or
edx:eax registers
 The idiv instruction performs the same
operations on (signed) values.
Signed Integer Division
 si_quotient = si_dividend /
si_divisor;
 1. mov eax, dword ptr [si_dividend]
 2. cdq
 3. idiv eax, dword ptr [si_divisor]
 4. mov dword ptr [si_quotient], eax
The cdq instruction copies the sign (bit 31) of the value in the eax
register into every bit position in the edx register.

NOTE: Assembly code generated by Visual C++

Unsigned Integer Division
 ui_quotient = ui1_dividend /
ui_divisor;
 5. mov eax, dword ptr [ui_dividend]
 6. xor edx, edx
 7. div eax, dword ptr [ui_divisor]
 8. mov dword ptr [ui_quotient], eax

NOTE: Assembly code generated by Visual C++

Error Detection
 The Intel division instructions div and idiv do not
set the overflow flag.
 A division error is generated if
 the source operand (divisor) is zero
 if the quotient is too large for the designated register
 A divide error results in a fault on interrupt vector 0.
 When a fault is reported, the processor restores the
machine state to the state before the beginning of
execution of the faulting instruction.
Vulnerabilities
 A vulnerability is a set of conditions that allows
violation of an explicit or implicit security policy.
 Security flaws can result from hardware-level integer
error conditions or from faulty logic involving
integers.
 These security flaws can, when combined with other
conditions, contribute to a vulnerability.
Vulnerabilities Section Agenda
 Integer overflow
 Sign error

 Truncation

 Non-exceptional
JPEG Example
 Based on a real-world vulnerability in the handling of
the comment field in JPEG files
 Comment field includes a two-byte length field
indicating the length of the comment, including the
two-byte length field.
 To determine the length of the comment string (for
memory allocation), the function reads the value in
the length field and subtracts two.
 The function then allocates the length of the
comment plus one byte for the terminating null byte.
Integer Overflow Example
 1. void getComment(unsigned int len, char *src) {
 2. unsigned int size;
0 byte malloc() succeeds
 3. size = len - 2;
 4. char *comment = (char *)malloc(size + 1);
 5. memcpy(comment, src, size);
 6. return;
Size is interpreted as a large
 7. } positive value of 0xffffffff
 8. int _tmain(int argc, _TCHAR* argv[]) {
 9. getComment(1, "Comment ");
 10. return 0;
 11. }
Possible to cause an overflow by creating
an image with a comment length field of 1
Memory Allocation Example
 Integer overflow can occur in calloc() and other
memory allocation functions when computing the
size of a memory region.
 A buffer smaller than the requested size is returned,
possibly resulting in a subsequent buffer overflow.
 The following code fragments may lead to
vulnerabilities:
 C: p = calloc(sizeof(element_t), count);
 C++: p = new ElementType[count];
Memory Allocation
 The calloc() library call accepts two
arguments
 the storage size of the element type
 the number of elements
 The element type size is not specified
explicitly in the case of new operator in C++.
 To compute the size of the memory required,
the storage size is multiplied by the number
of elements.
Overflow Condition
 Ifthe result cannot be represented in a
signed integer, the allocation routine can
appear to succeed but allocate an area that is
too small.
 The application can write beyond the end of

the allocated buffer resulting in a heap-based

buffer overflow.
 Program accepts two
arguments (the length
Sign Error Example 1 of data to copy and
the actual data)
 1. #define BUFF_SIZE 10
 2. int main(int argc, char* argv[]){
 3. int len; len declared as a signed integer
 4. char buf[BUFF_SIZE];

argv[1] can be
5. len = atoi(argv[1]); a negative value
 6. if (len < BUFF_SIZE){
A negative
 7. memcpy(buf, argv[2], len); value
 8. } bypasses
 the check
9. }
Value is interpreted as an
unsigned value of type size_t
Sign Errors Example 2
 The negative length is interpreted as a large,
positive integer with the resulting buffer
overflow
 This vulnerability can be prevented by
restricting the integer len to a valid value
 more effective range check that guarantees len is
greater than 0 but less than BUFF_SIZE
 declare as an unsigned integer
 eliminates the conversion from a signed to unsigned
type in the call to memcpy()
 prevents the sign error from occurring
Truncation:
Vulnerable Implementation
 1. bool func(char *name, long cbBuf) {
 2. unsigned short bufSize = cbBuf;
 3. char *buf = (char *)malloc(bufSize);
 4. if (buf) {
 5. memcpy(buf, name, cbBuf);
 6. if (buf) free(buf); cbBuf is used to initialize
 7. return true; bufSize which is used
to allocate memory for
 8. } buf
 9. return false;
 10. } cbBuf is declared as a long and used
as the size in the memcpy() operation
Vulnerability 1
 cbBuf is temporarily stored in the unsigned short
bufSize.
 The maximum size of an unsigned short for both
GCC and the Visual C++ compiler on IA-32 is
65,535.
 The maximum value for a signed long on the
same platform is 2,147,483,647.
 A truncation error will occur on line 2 for any values
of cbBuf between 65,535 and 2,147,483,647.
Vulnerability 2
 This would only be an error and not a
vulnerability if bufSize were used for both
the calls to malloc() and memcpy()
 Because bufSize is used to allocate the

size of the buffer and cbBuf is used as the

size on the call to memcpy() it is possible to
overflow buf by anywhere from 1 to
2,147,418,112 (2,147,483,647 - 65,535)
bytes.
Non-Exceptional Integer Errors
 Integer related errors can occur without an
exceptional condition (such as an overflow)
occurring
Negative Indices
 1. int *table = NULL;\
 2. int insert_in_table(int pos, int value){
 3. if (!table) {
 4. table = (int *)malloc(sizeof(int) * 100);
 5. }
 Storage for the
6. if (pos > 99) {

array is
7. return -1;
allocated on
 8. } pos is not > 99
the heap
 9. table[pos] = value;
 10. return 0;
 11. }
value is inserted into the
array at the specified position
Vulnerability
 There is a vulnerability resulting from
incorrect range checking of pos
 Because pos is declared as a signed integer,
both positive and negative values can be passed
to the function.
 An out-of-range positive value would be caught
but a negative value would not.
Mitigation

 Type range checking

 Strong typing

 Compiler checks

 Safe integer operations

 Testing and reviews

Type Range Checking
 Type range checking can eliminate integer
vulnerabilities.
 Languages such as Pascal and Ada allow range
restrictions to be applied to any scalar type to form
subtypes.
 Ada allows range restrictions to be declared on
derived types using the range keyword:
type day is new INTEGER range 1..31;
 Range restrictions are enforced by the language
runtime.
 C and C++ are not nearly as good at enforcing type
safety.
Type Range Checking Example
 1. #define BUFF_SIZE 10
 2. int main(int argc, char* argv[]){
 3. unsigned int len; Implicit type check from
 4. char buf[BUFF_SIZE]; the declaration as an
 5. len = atoi(argv[1]);  unsigned integer
.
 6. if ((0<len) && (len<BUFF_SIZE) ){
 7. memcpy(buf, argv[2], len);
 8. }
 9. else Explicit check for both upper and lower bounds
 10. printf("Too much data\n");
 11. }
Range Checking
 External inputs should be evaluated to determine
whether there are identifiable upper and lower
bounds.
 these limits should be enforced by the interface
 easier to find and correct input problems than it is to trace
internal errors back to faulty inputs
 Limit input of excessively large or small integers
 Typographic conventions can be used in code to
 distinguish constants from variables
 distinguish externally influenced variables from locally used
variables with well-defined ranges
Strong Typing
 One way to provide better type checking is to
provide better types.
 Using an unsigned type can guarantee that a

variable does not contain a negative value.

 This solution does not prevent overflow.

 Strong typing should be used so that the

compiler can be more effective in identifying

range problems.
Strong Typing Example
 Declare an integer to store the temperature of water
using the Fahrenheit scale
 unsigned char waterTemperature;
 waterTemperature is an unsigned 8-bit value in
the range 1-255
 unsigned char
 sufficient to represent liquid water temperatures which
range from 32 degrees Fahrenheit (freezing) to 212
degrees Fahrenheit (the boiling point).
 does not prevent overflow
 allows invalid values (e.g., 1-31 and 213-255).
Abstract Data Type
 One solution is to create an abstract data type in
which waterTemperature is private and cannot
be directly accessed by the user.
 A user of this data abstraction can only access,
update, or operate on this value through public
method calls.
 These methods must provide type safety by
ensuring that the value of the waterTemperature
does not leave the valid range.
 If implemented properly, there is no possibility of an
integer type range error occurring.
Visual C++ Compiler Checks
 Visual C++ .NET 2003 generates a warning
(C4244) when an integer value is assigned to a
smaller integer type.
 At level 1 a warning is issued if __int64 is assigned to
unsigned int.
 At level 3 and 4, a “possible loss of data” warning is issued
if an integer is converted to a smaller type.
 For example, the following assignment is flagged
at warning level 4
 int main() {
int b = 0, c = 0;
 short a = b + c; // C4244
}
Visual C++ Runtime Checks
 Visual C++ .NET 2003 includes runtime checks that
catch truncation errors as integers are assigned to
shorter variables that result in lost data.
 The /RTCc compiler flag catches those errors and
creates a report.
 Visual C++ includes a runtime_checks pragma
that disables or restores the /RTC settings, but does
not include flags for catching other runtime errors
such as overflows.
 Runtime error checks are not valid in a release
(optimized) build for performance reasons.
GCC Runtime Checks
 GCC compilers provide an -ftrapv option
 provides limited support for detecting integer
exceptions at runtime.
 generates traps for signed overflow for addition,
subtraction, and multiplication
 generates calls to existing library functions
 GCC runtime checks are based on post-
conditions—the operation is performed and
the results are checked for validity
Postcondition
 For unsigned integers if the sum is smaller than
either operand, an overflow has occurred
 For signed integers, let sum = lhs + rhs
 If lhs is non-negative and sum < rhs, an overflow has
occurred.
 If lhs is negative and sum > rhs, an overflow has
occurred.
 In all other cases, the addition operation succeeds
Adding Signed Integers
Function from the gcc runtime system used to detect errors
resulting from the addition of signed 16-bit integers

 1. Wtype __addvsi3 (Wtype a, Wtype b)

{
 2. const Wtype w = a + b;
 3. if (b >= 0 ? w < a : w > a)
 4. abort (); The addition is performed
 5. return w; and the sum is compared to
the operands to determine if
 6. } an error occurred
abort() is called if
• b is non-negative and w < a
• b is negative and w > a
Safe Integer Operations 1
 Integer operations can result in error conditions and
possible lost data.
 The first line of defense against integer
vulnerabilities should be range checking
 Explicitly
 Implicitly - through strong typing
 It is difficult to guarantee that multiple input variables
cannot be manipulated to cause an error to occur in
some operation somewhere in a program.
Safe Integer Operations 2
 An alternative or ancillary approach is to
protect each operation.
 This approach can be labor intensive and

expensive to perform.
 Use a safe integer library for all operations on

integers where one or more of the inputs

could be influenced by an untrusted source.
Safe Integer Solutions
C language compatible library
 Written by Michael Howard at Microsoft
 Detects integer overflow conditions using IA-32
specific mechanisms
Unsigned Add Function
 1. in bool UAdd(size_t a, size_t b, size_t *r)
{
 2. __asm {
 3. mov eax, dword ptr [a]
 4. add eax, dword ptr [b]
 5. mov ecx, dword ptr [r]
 6. mov dword ptr [ecx], eax
 7. jc short j1
 8. mov al, 1 // 1 is success
 9. jmp short j2
 10. j1:
 11. xor al, al // 0 is failure
 12. j2:
 13. };
 14. }
Unsigned Add Function
Example
 1. int main(int argc, char *const *argv) {
 2. unsigned int total;
 3. if (UAdd(strlen(argv[1]), 1, &total) &&
 UAdd(total, strlen(argv[2]), &total)) {
 4. char *buff = (char *)malloc(total);
 5. strcpy(buff, argv[1]);
 6. strcat(buff, argv[2]);
 7. else {
 8. abort();
 9. } The length of the combined strings is
 10. } calculated using UAdd() with appropriate
checks for error conditions.
SafeInt Class
 SafeInt is a C++ template class written by
David LeBlanc.
 Implements a precondition approach that
tests the values of operands before
performing an operation to determine if an
error will occur.
 The class is declared as a template, so it can
be used with any integer type.
 Every operator has been overridden except
for the subscript operator[]
 The variables s1 and s2 are
declared as SafeInt types
SafeInt Example
 1. int main(int argc, char *const *argv) {
 2. try{
 3. SafeInt<unsigned long> s1(strlen(argv[1]));
 4. SafeInt<unsigned long> s2(strlen(argv[2]));
 5. char *buff = (char *) malloc(s1 + s2 + 1);
 6. strcpy(buff, argv[1]);
 7. strcat(buff, argv[2]);
 8. }
 9. catch(SafeIntException err) {
 10. abort();
 11. }
 12. }
When the + operator is invoked it uses the
safe version of the operator implemented as
part of the SafeInt class.
Addition
 Additionof unsigned integers can result in an
integer overflow if the sum of the left-hand
side (LHS) and right-hand side (RHS) of an
addition operation is greater than
 UINT_MAX for addition of unsigned int type
 ULLONG_MAX for addition of unsigned long
long type
Safe Integer Solutions
Compared
 SafeInt library has several advantages
 more portable than safe arithmetic operations that
depend on assembly language instructions.
 more usable
 operators can be used inline in expressions
 SafeInt uses C++ exception handling
 better performance (with optimized code)
 Fails
to provide correct integer promotion
behavior
When to Use Safe Integers
 Use safe integers when integer values can be
manipulated by untrusted sources, for example
 the size of a structure
 the number of structures to allocate

 void* CreateStructs(int StructSize, int HowMany)

{
 SafeInt<unsigned long> s(StructSize);
 s *= HowMany; Structure size multiplied by # required to
determine size of memory to allocate.
 return malloc(s.Value());
 }

The multiplication can overflow the integer and create a

buffer overflow vulnerability
When Not to Use Safe Integers
 Don’t use safe integers when no overflow possible
 
tight loop
 variables are not externally influenced
 void foo() {
 char a[INT_MAX];
 int i;
 for (i = 0; i < INT_MAX; i++)
 a[i] = '\0';
 }
Testing 1
 Input validation does not guarantee that
subsequent operations on integers will not
result in an overflow or other error condition.
 Testing does not provide any guarantees

either
 It is impossible to cover all ranges of possible
inputs on anything but the most trivial programs.
 If applied correctly, testing can increase
confidence that the code is secure.
Testing 2
 Integer vulnerability tests should include boundary
conditions for all integer variables.
 If type range checks are inserted in the code, test that they
function correctly for upper and lower bounds.
 If boundary tests have not been included, test for minimum
and maximum integer values for the various integer sizes
used.
 Use white box testing to determine the types of
integer variables.
 If source code is not available, run tests with the
various maximum and minimum values for each
type.
Source Code Audit
 Source code should be audited or inspected for
possible integer range errors
 When auditing, check for the following:
 Integer type ranges are properly checked.
 Input values are restricted to a valid range based on their
intended use.
 Integers that do not require negative values are
declared as unsigned and properly range-checked
for upper and lower bounds.
 Operations on integers originating from untrusted
sources are performed using a safe integer library.
Notable Vulnerabilities
 Integer Overflow In XDR Library
 SunRPC xdr_array buffer overflow
 http://www.iss.net/security_center/static/9170.php
 Windows DirectX MIDI Library
 eEye Digital Security advisory AD20030723
 http://www.eeye.com/html/Research/Advisories/
AD20030723.html
 Bash
 CERT Advisory CA-1996-22
 http://www.cert.org/advisories/CA-1996-22.html