TU Wien:Introduction to Security VU (Maffei)/Prüfung 2021-06-25

CSRF

Which of the following can be used as a protection or mitigation mechanism against CSRF attacks?

1. The HttpOnly cookie attribute
2. The validation of the Referer HTTP header.
3. The SameSite cookie attribute
4. The web application embeds secret, randomly generated tokens into the HTML forms to be protected. When a request is received, the application checks for the presence of the token before processing it.
5. The Content Security Policy.
6. The Secure cookie attribute

Cookie jar

The cookie jar of a browser contains the following cookies:

Name	Value	Domain attribute	Domain who set the cookie	Path	Secure	HttpOnly	SameSite
sid	xyz123	not set	example.com	/	yes	yes	Lax
lang	en	example.com	prefs.example.com	/	no	no	Lax
admin	abc456	not set	example.com	/admin	yes	yes	Strict

Which cookies are attached to a request from https://site.com to https://www.example.com/admin/index.php? Assume that the request been triggered by the user clicking on a link on https://site.com, i.e., it is a top-level navigation.

1. lang
2. sid
3. admin

User’s Browser

Which of the following web attacks involve the user’s browser as part of the attack flow?

1. Command Injection
2. Reflected Cross-Site Scripting
3. Path Traversal
4. Stored Cross-Site Scripting
5. Cross-Site Request Forgery

ECB

In the Electronic Codebook (ECB) encryption mode…

1. Decryption cannot be parallelized.
2. Identical blocks in the plaintext result in identical blocks in the ciphertext.
3. If one bit in the first block of the ciphertext gets corrupted, none of the plaintext blocks can be recovered.
4. Encryption cannot be parallelized.
5. Random read access is possible.

Prepared Statements

Which claims about prepared statements are correct (assuming a correct usage)?

1. Prepared statements are more effective if malicious characters are filtered out before their usage.
2. Prepared statements are used to prevent stored XSS.
3. Prepared statements remove all special characters from user input.
4. Prepared statements guarantee the structural integrity of a SQL query.

iFrame

The page at https://www.example.com/admin/index.php contains a frame whose source attribute can be chosen among the URLs below. Assume that the frame executes a script which tries to access the DOM of the page embedding the frame. For which URLs does the operation succeed?

1. https://www.example.com:443/index.php
2. https://example.com/list.php?name=matteo
3. http://www.example.com/admin/index.php
4. https://www.example.com/list.php?name=karl

Protocol

Consider the following protocol, where we use

• skA to denote A’s private signing key,
• pkB to denote B’s public encryption key,
• sign(m;k) for a message m signed with the private signing key k,
• enc(m;k) for a public key encryption of message m with public key k.
• n for a nonce.

A                                       B

               <— A,B,n —

      —  enc(m,sign(n,m;skA);pkB) —>

Which of the following statements about the protocol are true?

1. The protocol ensures the confidentiality of the message m.
2. The protocol provides non-injective agreement: whenever B authenticates m as coming from A, A has really started an session with B to authenticate m). However, it does not provide injective agreement.
3. The protocols provides injective agreement: whenever B authenticates the message m as coming from A, A has really started a session with B to authenticate m and the authentication request is fresh.

XSS

Which of the following claims about XSS are true?

1. XSS attacks are not possible if the HTTPS protocol is used.
2. Cookies marked as HttpOnly cannot be read from the cookie jar in a XSS attack.
3. It is sufficient to remove <script> tags from user input to prevent all XSS attacks.
4. Reflected XSS attacks are impossible if a webpage does not use GET-parameters.

2-factor authentication

Which of the following combinations can be used to implement 2-factor authentication (2FA)?

1. Password + key stretching
2. Password + authenticator app
3. Fingerprint + hardware token
4. OTP + CAPTCHA

Subsititution Cipher

The following message has been encrypted with a substitution cipher:

PBBDCT

Which of the following could be the corresponding plaintext message?

1. ALERTS
2. ACCEPT
3. ASSERT
4. ERRORS

Unix Permissions

Assume the following output of the ls command displaying the permissions of the standard rm binary:

-rwsr-x-w- 1 root users 72056 Sep  5  2019 /usr/bin/rm

Which of the following statements are true?

1. The user jane, who is not a member of the users group, can execute the rm binary
2. The root user can not execute the rm binary
3. The permissions allow unprivileged users (i.e., users other than root and members of the users group) to cause severe damage in the system
4. The user john in the users group can execute rm to delete any file owned by root

Stack canaries

Which of the following statements about stack canaries are correct?

1. Stack canaries are an effective mitigation against heap overflow vulnerabilities
2. Given this vulnerable function call in C memcpy(dest, src, strlen(src)+1), where the attacker controls dest and the content of src, stack canaries prevent attackers from overwriting the return address of the current function
3. Stack canaries do not prevent attackers from overwriting variables in the same stack frame
4. The protection mechanism is voided if an attacker overwrites the stack location containing the canary value

PHP

Consider the following snippet of code:

<?php
    $db = new PDO(CONNECTION_STRING, DB_USER, DB_PASS);

    $query = "SELECT name, mail_address FROM users WHERE POSITION(?, name) > 0";

    $sth = $db->prepare($query);
    $sth->bindValue(1, $_GET["search"]);
    $sth->execute();
    
    echo "Results for " . $_GET["search"];

    foreach ($sth as $row) {
        echo "Name: " . $row["name"];
        echo "Mail address: " . $row["mail_address"];
    }
?>

Which of the following statements are true?

1. The code is vulnerable to command injection.
2. The code is secure.
3. The code is vulnerable to SQL injection.
4. The code is vulnerable to reflected XSS.
5. The code is vulnerable to path traversal attacks.

RSA & OEAP

Consider an encryption scheme based on RSA with OEAP padding. We write E(m, k) to denote the encryption of message m under the public key k = (e, N). Which of the following statements are true?

1. Given E(m, k) where the message m is much smaller than the modulus N and e=3, it is easy to recover m.
2. The multiplicative property of RSA holds, i.e., E(m1, k) * E(m2, k) = E(m1*m2, k) for any m1, m2.
3. OEAP introduces randomness in the ciphertext.

Bruteforce

Which techniques can be used to reduce the likelihood of a successful bruteforce attack against a login form on a website?

1. Password salting
2. Cascade hashing (~1M iterations) during the authentication phase
3. 2-factor authentication
4. CAPTCHA

Type system 1

Consider the following information flow type system presented during the lecture.

Which types of information leaks can be detected by it?

1. Explicit information flows, i.e., assignment of a secret (high) value to a public (low) variable.
2. Information flows via termination channels.
3. Implicit information flows where the value of a public variable is assigned within a conditional statement whose condition involves a secret variable.
4. Information flows via timing channels.

W^X

Which of the following statements about W^X/DEP are correct?

1. W^X/DEP prevents the stack from being corrupted to alter the control flow of the application
2. Code reuse attacks are not mitigated by W^X/DEP
3. W^X/DEP protects against code injection
4. W^X/DEP works by randomizing the writable memory

Declassification

Which of the following statements regarding declassification are correct?

1. A typical use case of declassification is the case of password checking, where the result of a comparison involving a secret (i.e., the password) is treated as a public value.
2. Declassification is used in information flow type systems to explicitly allow some form of information leaks in programs.
3. Improper usage of declassification may introduce security vulnerabilities in programs.

C

Consider the following C program:

#include <stdio.h>
#include <string.h>

int
valid_username(void) {
    char username[16];
    gets(username);
    if(strlen(username) > 15) {
        return 0;
    } else {
        return 1;
    }
}

int
main(void) {
    if(valid_username()) {
        printf("All fine!\n");
    };

    return 0;
}

Which of the following statements are correct?

1. The input “Markchristopher” does not cause a buffer overflow
2. The correct way to enforce bound-checking is to truncate the username array after the gets call
3. The fundamental problem is the limited array length of the variable username: increasing the length solves the problem
4. The input “supercalifragilisticexpialidocious” does not cause a buffer overflow because the user input is ensured to be shorter than 16 bytes

Type system 2

Consider the information flow type system presented during the lecture, shown in the picture below, where we let n range over integers and booleans and e over arithmetic and boolean expressions.

Assume that variables whose name start with the letter h are classified as high (H, secret), while the others are considered as low (L, public). When typing, an expression e cannot be assigned type L if any of its operands has type H.

Which of the following programs would successfully typecheck?

1)

if (l1 < l2)
  h := true
else
  h := false

2)

h := 5

3)

l := 0;
while (h >= l) l := l+1

4)

l := h1 + h2;
l := 0

5)

if (h > 0)
  if (l1 > l2)
    la = l1 - l2
  else
    la = l2 - l1
else
  skip

6)

if (h1 > h2)
  hmin := h2
else
  hmin := h1

Note: Solutions can be found at [1]