CVE-2016-10192

aka Streaming Worm

CVE: CVE-2016-10192
CWE: 119
ipc:
  note: 'Shortly after the mitigation, the ffserver was refactored out of ffmpeg.

    '
  answer: false
  question: |
    Did the feature that this vulnerability affected use inter-process
    communication? IPC includes OS signals, pipes, stdin/stdout, message
    passing, and clipboard. Writing to files that another program in this
    software system reads is another form of IPC.

    Answer should be boolean.
CVSS: AV:N/AC:L/Au:N/C:P/I:P/A:P
bugs:
- 'ffserver: Segmentation fault when streaming RTSP with ffserver (Always)'
- ffserver crash when GET sdp file
i18n:
  note: 'There was no evidence of i18n being applicable to this vulnerabilty.

    '
  answer: false
  question: |
    Was the feature impacted by this vulnerability about internationalization
    (i18n)? An internationalization feature is one that enables people from all
    over the world to use the system. This includes translations, locales,
    typography, unicode, or various other features.

    Answer should be boolean. Write a note about how you came to the conclusions
    you did.
vccs:
- note: "The commit in which the ffserver.c file is updated to work with chunked \nencoding
    HTTP streams, which causes the buffer overflow. Identified by \narcheogit and
    verified by Fahd Masood\n"
  commit: 19c8c4ecf4e26762327a472dcf34132ceb196e30
- note: ''
  commit: ''
fixes:
- note: |
    Fix in most recent version. An assertion was used to check the length of chunk_size
    and ensure it wasn't negaitve, which ultimately mitigated the vulnerabilty.
  commit: a5d25faa3f4b18dac737fdb35d0dd68eb0dc2156
- note: Fix for version 3.1.6
  commit: 37904d11779482f375b13da24f33f75daf13638f
- note: Fix for version 3.0.5
  commit: 1768e02a046ac05cb212991ae23021ad412cd15a
- note: Fix for version 2.8.10
  commit: e0cb113f9b4b7a26ac0053a483f92c26a4a90f0e
- note: Fix for version 2.4.14
  commit: a07cf6423ec30be54493d46ad541c8f91fe1c13e
bounty:
  amt: 
  url: 
  announced: 
lessons:
  yagni:
    note: 
    applies: 
  question: |
    Are there any common lessons we have learned from class that apply to this
    vulnerability? In other words, could this vulnerability serve as an example
    of one of those lessons?

    Leave "applies" blank or put false if you did not see that lesson (you do
    not need to put a reason). Put "true" if you feel the lesson applies and put
    a quick explanation of how it applies.

    Don't feel the need to claim that ALL of these apply, but it's pretty likely
    that one or two of them apply.

    If you think of another lesson we covered in class that applies here, feel
    free to give it a small name and add one in the same format as these.
  serial_killer:
    note: 
    applies: 
  complex_inputs:
    note: 
    applies: 
  distrust_input:
    note: The attacker can craft a video that causes heap-based buffer overflow
    applies: true
  least_privilege:
    note: 
    applies: 
  native_wrappers:
    note: 
    applies: 
  defense_in_depth:
    note: 
    applies: 
  secure_by_default:
    note: 
    applies: 
  environment_variables:
    note: 
    applies: 
  security_by_obscurity:
    note: 
    applies: 
  frameworks_are_optional:
    note: 
    applies: 
upvotes: 6
CWE_note: 
lifetime:
  answer: "After inspecting the commit messages between VCC and the fix, it was clear
    that\nmultiple developers were working on this part of the program. Many of the
    commit \nmessages were unclear, which probably raised confusion amongst \nthe
    team.\n"
  question: |
    We consider the "lifetime" of this vulnerability to be from the earliest
    VCC to the latest fix.

    How would you characterize the development of this vulnerable code during
    its lifetime? Many new features? Neglected? One developer?
    Massive refactoring? Changing big dependencies?
    Changing architectures or language?


    What about the time period for the project itself?
    Did it undergo big changes during this time?

    Look at:
      * the commit messages between the two, using a command like
          git log --stat abc..def -- your/file.c
          (where a)
      * the changelog between the two
mistakes:
  answer: "A lack of understanding was responsible for the vulnerabilty being introduced\nin
    the first place. The minimal and lack thereof checking of bounds on input was
    \na serious implementation flaw that created havoc amongst the developers. \nThe
    bug reports suggested that the programers were pretty clueless in their \nattempts
    to navigate this vulnerabilty (i.e. seg_faults everywhere). \nThis could have
    potentially been due to a lack of design specifications, which \nled the developers
    to code freely and not write secure code at every turn \n(i.e. canory style bounds
    checking not used until mitigation). Communication \nissues and too many developers
    working on this code-base could have resulted \nin the 6 year wait for this mitigation
    (2010 to 2016). Additionally, the ffmpeg \nserver was out-of-scope, which ultimately
    led to it being refactored out of the \nlatest product versions.\n"
  question: |
    In your opinion, after all of this research, what mistakes were made that
    led to this vulnerability? Design mistakes? Maintainability? Requirements?
    Miscommunications? Lack of testing? Lack of understanding? Lack of
    specifications? Working alone?

    An answer like "it was just a coding mistake" is not thoughtful enough.
    If it's such an easy mistake to make, how was it missed?

    Also, look at the CWE entry for this vulnerability and examine the mitigations
    they have written there. Are they doing those? Does the fix look proper?

    Use those questions to inspire your answer. Don't feel obligated to answer
    every one. Write a thoughtful entry here that those ing the software
    engineering industry would find interesting.
nickname: Streaming Worm
subsystem:
  name: libavcodec
  answer: "The vulnerabilty was discovered in the libavcodec subsystem. The subsystem
    was \ndiscovered by going into the VCCS history and finding the original file,
    in which\na couple of gcc statements showed the which subsystem it belonged to.\n"
  question: |
    What subsystems was the mistake in?

    Look at the path of the source code files code that were fixed to get
    directory names. Look at comments in the code. Look at the bug reports how
    the bug report was tagged.

    In FFmpeg, the Component field is useful here. Often people will say
      "undetermined" - do more homework than them and make a judgement call if
      you can.

    Note: a filepath is NOT a subsystem - that's too granular.

    A non-exhaustive examples for ffmpeg are:
      * avcodec
      * avdevice
      * avfilter
      * avformat
      * avutil
      * build system
      * documentation
      * fate
      * ffplay
      * ffprobe
      * postproc
      * swresample
      * swscale
      * tools
      * trac
      * website

    If this involves fixing multiple subsystems, you can make this an array.

    In 'answer', explain how you arrived at this determination.
discovered:
  answer: "There were numerous bug reports found in the ffmeg database that indicated
    a \nsegmentation fault in ffserver.c. The Heap-based buffer overflow vulnerability
    \nwas found by a developer, Paul Cher and signed-off by his supervisor, Michael
    \nNiedermayer.\n"
  contest: false
  question: |
    How was this vulnerability discovered?

    Find any relevant bug reports or mailing list conversations read the
    conversation to find out how this was
    originally found. Answer in longform below in "answer".

    If it's clear that the vulenrability was discovered by a
    contest, fill in the name of the contest in "answer".

    If it's clear that it was found by a developer of the project itself,
    e.g. from the email address or a previous committer.

    The automated, contest, and developer flags can be true, false, or nil.

    If there is no evidence as to how this vulnerability was found, then please
    explain where you looked. Thus, 'answer' should always have some
    explanation.
  automated: false
  developer: true
description: "There was a Heap-based buffer overflow vulnerability in ffmpeg's implementation
  of a \nmultimedia streaming server for live broadcasts that allows attackers to
  execute \narbitrary code by leveraging failure to check chunk size. \n\nInitial
  inception of the vulnerability was in 1988 and was nicknamed the \nMorris Internet
  worm. In ffserver.c, the buffer overflow was caused by not \nchecking for a negative
  chunk_size, which immediately resulted in rip control \nand then code execution\n"
unit_tested:
  fix: false
  code: true
  question: |
    Were automated unit tests involved in this vulnerability?
    Was the original code unit tested, or not unit tested? Did the fix involve
    improving the automated tests?

    For code: and fix: - your answer should be boolean.

    For the code_answer below, look not only at the fix but the surrounding
    code near the fix in related directories and determine if and was there were
    unit tests involved for this subsystem. The code

    For the fix_answer below, check if the fix for the vulnerability involves
    adding or improving an automated test to ensure this doesn't happen again.

    In FFmpeg, they have the FATE suite of regression tests. They are not
    "unit" tests per se, but are automated regression tests. If the fix for the
    vulnerability involved adding or updating a FATE suite, then make fix: true.
  fix_answer: No additional unit tests were added upon the mitigation of buffer overflow
  code_answer: There were unit tests for this tool in ffserver.config
reported_date: '2016-12-05'
specification:
  answer: false
  answer_note: No violation was found in the commits/bug reports
  instructions: |
    Is there mention of a violation of a specification? For example,
    an RFC specification, a protocol specification, codec spec, or a requirements
    specification.

    Be sure to check all artifacts for this: bug report, security
    advisory, commit message, etc.

    The answer field should be boolean. In answer_note, please explain
    why you come to that conclusion.
announced_date: 2017-02-09T15:59Z
curation_level: 1.0
published_date: 2017-02-09T15:59Z
CWE_instructions: |
  Please go to http://cwe.mitre.org and find the most specific, appropriate CWE
  entry that describes your vulnerability. We recommend going to
  https://cwe.mitre.org/data/definitions/699.html for the Software Development
  view of the vulnerabilities. We also recommend the tool
  http://www.cwevis.org/viz to help see how the classifications work.

  If you have anything to note about why you classified it this way, write
  something in CWE_note. This field is optional.

  Just the number here is fine. No need for name or CWE prefix. If more than one
  apply here, then choose the best one and mention the others in CWE_note.
autodiscoverable:
  answer: true
  answer_note: "A fuzzer can be implemented to mitigate buffer overflows by targeting
    bug \ncandidates and use a novel “spot-check” approach in symbolic execution.\n"
  instructions: |
    Is it plausible that a fully automated tool could have discovered
    this? These are tools that require little knowledge of the domain,
     e.g. automatic static analysis, compiler warnings, fuzzers.

    Examples for true answers: SQL injection, XSS, buffer overflow,
    use-after-free

    Examples for false: specification violations, permissions issues, anything
    that would require a tool to be "aware" of the project's
    domain-specific requirements.

    The answer field should be boolean. In answer_note, please explain
    why you come to that conclusion.
bugs_instructions: |
  Are there any bug reports that trace to this vulnerability?

  Sometimes a commit message will mention a bug, or the security page on FFmpeg.
  Sometimes you need to search bug database, which is here:
      https://trac.ffmpeg.org
yaml_instructions: |
  ===YAML Primer===
  This is a dictionary data structure, akin to JSON.
  Everything before a colon is a key, and the values here are usually strings

  For one-line strings, you can just use quotes after the colon

  For multi-line strings, as we do for our instructions, you put a | and then
  indent by two spaces.

  For readability, we hard-wrap multi-line strings at 80 characters. This is
  not absolutely required, but appreciated.
  === End YAML Primer===
bounty_instructions: |
  If you came across any indications that a bounty was paid out for this
  vulnerability, fill it out here. Or correct it if the information already here
  was wrong. Otherwise, leave it blank.
interesting_commits:
  commits:
  - note: "On July 10th 2016, FFMPEG decided to remove the ffserver tool from their
      project\ndue to it being problematic to mantain because of its use of internal
      APIs,\nwhich complicated their at-the-time cleanups to the libavformat library,
      \nand block further cleanups and improvements which were desired by API users.\n"
    commit: da38da459598c0a8777522720b6a4afd433e29bc
  - note: "On the attempt to play a stream on a remote PC, ffserver terminates with
      a \nseg_fault.\n"
    commit: 9200514ad8717c63f82101dc394f4378854325bf
  question: |
    Optional: are there any interesting commits between your VCC(s) and fix(es)?

    Write a brief (under 100 words) description of why you think this commit was
    interesting in light of the lessons learned from this vulnerability. Any
    emerging themes?

    This is a catch-all for any interesting commit you find.
    Example of interesting commits:
      * Giant refactors
      * Changing an API or big dependency
      * A change that looks really sketchy
      * Commits where a developer mentions security in their message or comments
curated_instructions: |
  If you are manually editing this file, then you are "curating" it.
  This will enable additional integrity checks on this file to make sure you
  fill everything out properly.

  IMPORTANT: If you are a student, we cannot accept your work as finished unless
  curation_level is properly set.

  The latest curation level is 1.0.
  If you are curating this vulnerability, set it to the latest curation level.
upvotes_instructions: |
  For the first round, ignore this upvotes number.

  For the second round of reviewing, you will be giving a certain amount of
  upvotes to each vulnerability you see. Your peers will tell you how
  interesting they think this vulnerability is, and you'll add that to the
  upvotes score on your branch.
nickname_instructions: |
  A catchy name for this vulnerability that would draw attention it. If the
  report mentions a nickname, e.g. "Heartbleed", use that. Or come up with one!

  Must be under 30 characters. Optional. Be appropriate.
reported_instructions: |
  What date was the vulnerability reported to the security team? Look at the
  security bulletins, bug reports, commit messages of the fix.

  Reported date is likely the same as announced and published. Leave blank if
  you can't find it out.

  Please enter your date in YYYY-MM-DD format.
announced_instructions: |
  Was there a date that this vulnerability was announced to the world? You can
  find this in changelogs, blogs, bug reports, or perhaps the CVE date.

  A good source of this for FFmpeg are their version numbers traced to tags:

    https://ffmpeg.org/security.html
    https://github.com/FFmpeg/FFmpeg/releases


  Please enter your date in YYYY-MM-DD format.
fixes_vcc_instructions: |
  Please put the commit hash in "commit" field below (see my example in
  CVE-2011-3092.yml).

  Fixes and VCCs follow the same format - they are arrays of hashes.

  The notes field is optional - place anything that clarifies things or you
  find interesting in there.
published_instructions: |
  Is there a published fix or patch date for this vulnerability?
  Please enter your date in YYYY-MM-DD format.
description_instructions: |
  You can get an initial description from the CVE entry on cve.mitre.org. These
  descriptions are a fine start, but they can be kind of jargony.

  Rewrite this description IN YOUR OWN WORDS. Make it interesting and easy to
  read to anyone with some programming experience. We can always pull up the NVD
  description later to get more technical.

  Try to still be specific in your description, but remove project-specific
  stuff. Remove references to versions, specific filenames, and other jargon
  that outsiders to this project would not understand. Technology like "regular
  expressions" is fine, and security phrases like "invalid write" are fine to
  keep too.

  Your target audience is people just like you before you took any course in
  security