SQLite should never crash, overflow a buffer, leak memory, or exhibit any other harmful behavior, even when presented with maliciously malformed SQL inputs or database files. SQLite should always detect erroneous inputs and raise an error, not crash or corrupt memory. Any malfunction caused by an SQL input or database file is considered a serious bug and will be promptly addressed when brought to the attention of the SQLite developers. SQLite is extensively fuzz-tested to help ensure that it is resistant to these kinds of errors.
Nevertheless, bugs happen. If you are writing an application that sends untrusted SQL inputs or database files to SQLite, there are additional steps you can take to help reduce the attack surface and prevent zero-day exploits caused by undetected bugs.
Applications that accept untrusted SQL inputs should take the following precautions:
Set the SQLITE_DBCONFIG_DEFENSIVE flag. This prevents ordinary SQL statements from corrupting the database file.
Consider using the sqlite3_set_authorizer() interface to limit the scope of SQL that will be processed.
Applications that accept untrusted database files should do the following:
Run PRAGMA integrity_check or PRAGMA quick_check on the database as the first SQL statement after opening the database files and prior to running any other SQL statements. Reject and refuse to process any database file containing errors.
Enable the PRAGMA cell_size_check=ON setting.
Do not enable memory-mapped I/O. In other words, make sure that PRAGMA mmap_size=0.
The precautions above are not required in order to use SQLite safely with potentially hostile inputs. However, they do provide an extra layer of defense against zero-day exploits and are encouraged for applications that pass data from untrusted sources into SQLite.