CVE-2022-35737 is a vulnerability in SQLite, specifically in versions 1.0.12 through 3.39.x before 3.39.2. It allows for an array-bounds overflow if a string argument to a C API contains billions of bytes. This vulnerability can lead to various consequences, ranging from a simple application crash to arbitrary code execution.

• The issue lies in the sqlite3_snprintf function’s code, used in C/C++ programming for database interaction. Passing an exceedingly large string input (over 2 GB) to this function can cause a crash, enabling a denial of service (DoS) attack.
  
• This bug likely entered the code 22 years ago and remained undetected due to the improbability of passing gigabytes of data as function parameters at that time.
  
• The vulnerability was closed with the release of SQLite 3.39.2 in July 2022. However, software that includes SQLite needs to be updated to incorporate this fix.
  

The vector string for this score is CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H, indicating network access for the attack, low attack complexity, no privileges required, no user interaction required, no scope change, and high impact on availability.

Affected Platforms:

• This vulnerability affects various platforms, including those running Amazon Linux, where updates have been provided to address this issue.
  

Mitigation and Impact:

• The vulnerability is specific to the interface for C applications and only if the code is compiled with certain parameters.
  
• The practical exploitability of this vulnerability is still a subject of research, with limitations on its impact based on how SQLite is compiled and used in applications.
  

Additional Considerations:

• This vulnerability is notable for its long presence in the SQLite code and the challenges in detecting it using standard testing methods like fuzzing.
  

In terms of exploitation, the vulnerability’s practical use for an attack is still under investigation, with researchers pointing to several limitations that reduce the likelihood of successful exploitation.