Homework 5: Exam Review




Rate this product

Intro to Computer Security Homework 5: Exam Review
Homework 5: Exam Review

Concisely answer the following questions. Limit yourself to at most 80 words per subquestion.
1. HTTPS. A self-signed certificate makes the claim that a public key belongs to a particular
server, without any trusted certificate authority (CA) to verify it. Browsers display a warning
message when a site presents such a certificate, but users often override these warnings.
Some websites use self-signed certs to avoid the trouble of obtaining a cert from a trusted
(a) Briefly explain how using HTTPS with a self-signed certificate provides protection
against a passive eavesdropper.
(b) How might a man-in-the-middle attacker compromise a site that uses a self-signed
certificate, assuming that the client ignores browser certificate warnings?
(c) Some sites use HTTPS with a certificate signed by a trusted CA for their login pages,
then set a session cookie and use HTTP for the other pages on the site. Briefly compare
the security of this design to the use, for all pages on the site, of:
i. a self-signed certificate;
ii. a certificate signed by a trusted CA.
2. Web attacks. Consider a fictitious social networking site called MyPlace. MyPlace has
millions of users, not all of whom are particularly security-conscious. To protect them, all
pages on the site use HTTPS.
(a) MyPlace’s homepage has a “Delete account” link which leads to the following page:
<pAre you sure you want to delete your account?</p
<form action=”/deleteuser” method=”post”
<input type=”hidden” name=”user” value=”{{username}}”</input
<input type=”submit” value=”Yes, please delete my account”</input
(The web server replaces {{username}} with the username of the logged-in user.)
The implementation of /deleteuser is given by the following pseudocode:
if account_exists(request.parameters[‘user’]):
return ‘<pThanks for trying MyPlace!</p’
return ‘<pSorry, ‘ + request.parameters[‘user’] + ‘, an error occurred.</p’
Assume that the attacker knows the username of an intended victim. What’s a simple
way that the attacker can exploit this design to delete the victim’s account without any
direct contact with the victim or the victim’s browser?
(b) Suppose that /deleteuser is modified as follows:
if validate_user_login_cookie(request.parameters[‘user’], request.cookies[‘login_cookie’]):
return ‘<pThanks for trying MyPlace!</p’
return ‘<pSorry, ‘ + request.parameters[‘user’] + ‘, an error occurred.</p’
where validate_login_cookie() checks that the cookie sent by the browser is authentic
and was issued to the specified username. Assume that login_cookie is tied to the user’s
account and difficult to guess.)
Despite these changes, how can the attacker use CSRF to delete the victim’s account?
(c) Suppose that the HTML form in (a) is modified to include the current user’s login_cookie
as a hidden parameter, and /deleteuser is modified like this:
if request.parameters[‘login_cookie’] == request.cookies[‘login_cookie’] and
validate_login_cookie(request.parameters[‘user’], request.cookies[‘login_cookie’]):
return ‘<pThanks for trying MyPlace!</p’
return ‘<pSorry, ‘ + request.parameters[‘user’] + ‘, an error occurred.</p’
The attacker can still use XSS to delete the victim’s account. Briefly explain how.
3. Secure programming. StackGuard is a mechanism for defending C programs against
stack-based buffer overflows. It detects memory corruption using a canary, a known value
stored in each function’s stack frame immediately before the return address. Before a function returns, it verifies that its canary value hasn’t changed; if it has, the program halts with
an error.
(a) In some implementations, the canary value is a 64-bit integer that is randomly generated
each time the program runs. Explain why this prevents the basic form of stack-based
buffer overflow attack discussed in lecture.
(b) What is a security drawback to choosing the canary value at compile time instead of at
run time? If the value must be fixed, why is 0 a particularly good choice?
(c) No matter how the canary is chosen, StackGuard cannot protect against all buffer overflow vulnerabilities. List two kinds of bugs that can corrupt the stack and allow the
adversary to take control, even with StackGuard in place.
4. Ethics. Consider the following scenario: A worm is infecting systems by exploiting a bug
in a popular server program. It is spreading rapidly, and systems where it is deleted quickly
become reinfected. A security researcher decides to launch a counterattack in the form of
a defensive worm. Whenever a break-in attempt comes from a remote host, the defensive
worm detects it, heads off the break-in, and exploits the same bug to spread to the attacking
host. On that host, it deletes the original worm. It then waits until that system is attacked,
and the cycle repeats.
(a) Many people would claim that launching such a counterattack in this scenario is ethically unacceptable. Briefly argue in support of this view.
(b) Are there circumstances or conditions under which an active security counterattack
would be ethically justified? Briefly explain your reasoning.
Submission Template
Submit your work via CTools. Make sure each answer is formatted as a single line, and that the
file you submit is in plain text format.
# Problem 1
1a. [Answer …]
1b. [Answer …]
1c. [Answer …]
# Problem 2
2a. [Answer …]
2b. [Answer …]
2c. [Answer …]
# Problem 3
3a. [Answer …]
3b. [Answer …]
3c. [Answer …]
# Problem 4
4a. [Answer …]
4b. [Answer …]