Assignment 3 – Due Tuesday, February 28

  1. Solve “Question 4” from the Formal Models for Cryptography reading.

  2. Joe Crypto is a little too confident in his ability to design secure cryptographic systems, and decides he will “fix” the two main problems with ECB mode: detectable patterns across repeated plaintext blocks, and deterministic encryption. He has some super-awesome code that performs AES encryption in ECB mode, and he wants to re-use that software, so he comes up with this algorithm: He is going to XOR blocks of plaintext with a randomly-initialized counter (i.e., a “nonce”), and then feed that output through his ECB mode encryption software. He has even decided to provide additional protection by encrypting the initial counter value, resulting in what he has named “Pre-CTR mode,” with encryption shown in the following diagram (this was adapted from the diagram for CTR mode encryption on the Wikipedia page for “Block cipher modes of operation”):

    1. Draw the corresponding decryption diagram, marking the part of the diagram that corresponds to a regular ECB-mode decryption (similar to how that’s shaded in the encryption diagram).

    2. Joe tests his new Pre-CTR Mode encryption with the classic “Tux image” to see if it hides patterns (unlike ECB mode). Here’s the comparison:

      Excellent! Problem, solved, right? Unfortunately, his newly-invented Pre-CTR mode is not even IND-CPA secure. Show this by clearly defining an attacker for the IND-CPA game, and proving that it works. (Hint: Carefully think through the solution to the previous question. You can define a similar adversary in this case, which will also never make any direct calls to the encryption oracle like in the previous question. It requires using plaintexts with two blocks as challenge ciphers, where one of them is very carefully chosen to “undo” the change that the plaintext counter XOR does, and will always win the INC-CPA game. You’re not restricted to this structure – if you need more than two blocks, that’s fine – but your adversary does need to work!)

  3. Consider two users on a Unix system, alice and bob. Both are members of the group develop. There is another user on the system, charlie, who is not a member of the develop group. Consider the following files in a directory that all users can access:

     -r--r-----  1 alice develop  3612 Oct 25 14:36 Makefile
 -rwxrwxr-x  1 alice develop  4823 Oct 27 11:04 execute
 -rw-rw-r--  1 alice alice    2145 Oct 27 11:31 notes.txt
 -rw-rw-r--  1 bob   develop  1098 Oct 26 21:56 source.c
 -rw-------  1 alice develop  2796 Oct 27 11:03 source.o
 -rwsrwxr-x  1 alice develop  4823 Oct 27 11:12 special

    For each question below, give a very brief explanation of your answer.

    1. Which files can alice modify?

    2. Which files can bob modify?

    3. Which files can charlie read?

    4. Which files can charlie modify?

    5. Programs named execute and special are the exact same program, but are two copies with different permissions. This program tries to open and modify the notes.txt file. If bob runs the version named execute, will the program be able to perform these operations?

    6. If bob runs the version named special, will the program be able to perform these operations?

    7. If execute and special perform differently for bob, why?

  4. The “capabilities” labtainer exercise is brief and can be completed in under 30 minutes, but it demonstrates a very important security feature of modern Linux systems. Complete the labtainer exercise, with the following change: you do not need to submit the lab report it refers to under “Submission,” nor do you have to answer the questions at the end of Task 2, but you still need to submit the .lab file which documents your work. Hint for Task 1.2: You may need to refer to the man page “cap_from_text” to see how to format the capability-setting command. Remember that any time you are working on a lab, you can always type checkwork in the main system window to see how many of the major requirements have been met (there are other things that are checked in grading, but this gives good high-level feedback).

    You should also answer this question as part of your written problem submission: What results did you see from Task 2? In other words, which open calls failed? Answer honestly: Did you correctly predict what was going to happen? Then explain in your own words why some calls failed and others succeeded.

  5. Complete the “file-deletion” labtainer exercise, including the lab report (which is submitted as part of the labtainer .lab file – make sure you save your report before you run the stoplab command). This labtainer explores some of the concepts regarding file deletion that we discuss in the “Physical Security” lecture. Note that the files in these labtainers are not just plain ASCII files, which they probably should be. They have (non-ASCII) quotation marks, which don’t appear that way when you do a hexdump (which only shows ASCII characters). If I could change the lab, I would, but I don’t control that.