[php] Are PDO prepared statements sufficient to prevent SQL injection?

The short answer is NO, PDO prepares will not defend you from all possible SQL-Injection attacks. For certain obscure edge-cases.

I'm adapting this answer to talk about PDO...

The long answer isn't so easy. It's based off an attack demonstrated here.

The Attack

So, let's start off by showing the attack...

$pdo->query('SET NAMES gbk');
$var = "\xbf\x27 OR 1=1 /*";
$query = 'SELECT * FROM test WHERE name = ? LIMIT 1';
$stmt = $pdo->prepare($query);
$stmt->execute(array($var));

In certain circumstances, that will return more than 1 row. Let's dissect what's going on here:

  1. Selecting a Character Set

    $pdo->query('SET NAMES gbk');
    

    For this attack to work, we need the encoding that the server's expecting on the connection both to encode ' as in ASCII i.e. 0x27 and to have some character whose final byte is an ASCII \ i.e. 0x5c. As it turns out, there are 5 such encodings supported in MySQL 5.6 by default: big5, cp932, gb2312, gbk and sjis. We'll select gbk here.

    Now, it's very important to note the use of SET NAMES here. This sets the character set ON THE SERVER. There is another way of doing it, but we'll get there soon enough.

  2. The Payload

    The payload we're going to use for this injection starts with the byte sequence 0xbf27. In gbk, that's an invalid multibyte character; in latin1, it's the string ¿'. Note that in latin1 and gbk, 0x27 on its own is a literal ' character.

    We have chosen this payload because, if we called addslashes() on it, we'd insert an ASCII \ i.e. 0x5c, before the ' character. So we'd wind up with 0xbf5c27, which in gbk is a two character sequence: 0xbf5c followed by 0x27. Or in other words, a valid character followed by an unescaped '. But we're not using addslashes(). So on to the next step...

  3. $stmt->execute()

    The important thing to realize here is that PDO by default does NOT do true prepared statements. It emulates them (for MySQL). Therefore, PDO internally builds the query string, calling mysql_real_escape_string() (the MySQL C API function) on each bound string value.

    The C API call to mysql_real_escape_string() differs from addslashes() in that it knows the connection character set. So it can perform the escaping properly for the character set that the server is expecting. However, up to this point, the client thinks that we're still using latin1 for the connection, because we never told it otherwise. We did tell the server we're using gbk, but the client still thinks it's latin1.

    Therefore the call to mysql_real_escape_string() inserts the backslash, and we have a free hanging ' character in our "escaped" content! In fact, if we were to look at $var in the gbk character set, we'd see:

    ?' OR 1=1 /*

    Which is exactly what the attack requires.

  4. The Query

    This part is just a formality, but here's the rendered query:

    SELECT * FROM test WHERE name = '?' OR 1=1 /*' LIMIT 1
    

Congratulations, you just successfully attacked a program using PDO Prepared Statements...

The Simple Fix

Now, it's worth noting that you can prevent this by disabling emulated prepared statements:

$pdo->setAttribute(PDO::ATTR_EMULATE_PREPARES, false);

This will usually result in a true prepared statement (i.e. the data being sent over in a separate packet from the query). However, be aware that PDO will silently fallback to emulating statements that MySQL can't prepare natively: those that it can are listed in the manual, but beware to select the appropriate server version).

The Correct Fix

The problem here is that we didn't call the C API's mysql_set_charset() instead of SET NAMES. If we did, we'd be fine provided we are using a MySQL release since 2006.

If you're using an earlier MySQL release, then a bug in mysql_real_escape_string() meant that invalid multibyte characters such as those in our payload were treated as single bytes for escaping purposes even if the client had been correctly informed of the connection encoding and so this attack would still succeed. The bug was fixed in MySQL 4.1.20, 5.0.22 and 5.1.11.

But the worst part is that PDO didn't expose the C API for mysql_set_charset() until 5.3.6, so in prior versions it cannot prevent this attack for every possible command! It's now exposed as a DSN parameter, which should be used instead of SET NAMES...

The Saving Grace

As we said at the outset, for this attack to work the database connection must be encoded using a vulnerable character set. utf8mb4 is not vulnerable and yet can support every Unicode character: so you could elect to use that instead—but it has only been available since MySQL 5.5.3. An alternative is utf8, which is also not vulnerable and can support the whole of the Unicode Basic Multilingual Plane.

Alternatively, you can enable the NO_BACKSLASH_ESCAPES SQL mode, which (amongst other things) alters the operation of mysql_real_escape_string(). With this mode enabled, 0x27 will be replaced with 0x2727 rather than 0x5c27 and thus the escaping process cannot create valid characters in any of the vulnerable encodings where they did not exist previously (i.e. 0xbf27 is still 0xbf27 etc.)—so the server will still reject the string as invalid. However, see @eggyal's answer for a different vulnerability that can arise from using this SQL mode (albeit not with PDO).

Safe Examples

The following examples are safe:

mysql_query('SET NAMES utf8');
$var = mysql_real_escape_string("\xbf\x27 OR 1=1 /*");
mysql_query("SELECT * FROM test WHERE name = '$var' LIMIT 1");

Because the server's expecting utf8...

mysql_set_charset('gbk');
$var = mysql_real_escape_string("\xbf\x27 OR 1=1 /*");
mysql_query("SELECT * FROM test WHERE name = '$var' LIMIT 1");

Because we've properly set the character set so the client and the server match.

$pdo->setAttribute(PDO::ATTR_EMULATE_PREPARES, false);
$pdo->query('SET NAMES gbk');
$stmt = $pdo->prepare('SELECT * FROM test WHERE name = ? LIMIT 1');
$stmt->execute(array("\xbf\x27 OR 1=1 /*"));

Because we've turned off emulated prepared statements.

$pdo = new PDO('mysql:host=localhost;dbname=testdb;charset=gbk', $user, $password);
$stmt = $pdo->prepare('SELECT * FROM test WHERE name = ? LIMIT 1');
$stmt->execute(array("\xbf\x27 OR 1=1 /*"));

Because we've set the character set properly.

$mysqli->query('SET NAMES gbk');
$stmt = $mysqli->prepare('SELECT * FROM test WHERE name = ? LIMIT 1');
$param = "\xbf\x27 OR 1=1 /*";
$stmt->bind_param('s', $param);
$stmt->execute();

Because MySQLi does true prepared statements all the time.

Wrapping Up

If you:

  • Use Modern Versions of MySQL (late 5.1, all 5.5, 5.6, etc) AND PDO's DSN charset parameter (in PHP = 5.3.6)

OR

  • Don't use a vulnerable character set for connection encoding (you only use utf8 / latin1 / ascii / etc)

OR

  • Enable NO_BACKSLASH_ESCAPES SQL mode

You're 100% safe.

Otherwise, you're vulnerable even though you're using PDO Prepared Statements...

Addendum

I've been slowly working on a patch to change the default to not emulate prepares for a future version of PHP. The problem that I'm running into is that a LOT of tests break when I do that. One problem is that emulated prepares will only throw syntax errors on execute, but true prepares will throw errors on prepare. So that can cause issues (and is part of the reason tests are borking).

Examples related to php

I am receiving warning in Facebook Application using PHP SDK Pass PDO prepared statement to variables Parse error: syntax error, unexpected [ Preg_match backtrack error Removing "http://" from a string How do I hide the PHP explode delimiter from submitted form results? Problems with installation of Google App Engine SDK for php in OS X Laravel 4 with Sentry 2 add user to a group on Registration php & mysql query not echoing in html with tags? How do I show a message in the foreach loop?

Examples related to security

Monitoring the Full Disclosure mailinglist Two Page Login with Spring Security 3.2.x How to prevent a browser from storing passwords JWT authentication for ASP.NET Web API How to use a client certificate to authenticate and authorize in a Web API Disable-web-security in Chrome 48+ When you use 'badidea' or 'thisisunsafe' to bypass a Chrome certificate/HSTS error, does it only apply for the current site? How does Content Security Policy (CSP) work? How to prevent Screen Capture in Android Default SecurityProtocol in .NET 4.5

Examples related to pdo

Pass PDO prepared statement to variables PDO::__construct(): Server sent charset (255) unknown to the client. Please, report to the developers Laravel Migration Error: Syntax error or access violation: 1071 Specified key was too long; max key length is 767 bytes PHP 7 RC3: How to install missing MySQL PDO PHP Connection failed: SQLSTATE[HY000] [2002] Connection refused ERROR: SQLSTATE[HY000] [2002] No connection could be made because the target machine actively refused it php artisan migrate throwing [PDO Exception] Could not find driver - Using Laravel How to check if a row exists in MySQL? (i.e. check if an email exists in MySQL) PDOException SQLSTATE[HY000] [2002] No such file or directory How do I configure php to enable pdo and include mysqli on CentOS?

Examples related to sql-injection

Preventing SQL injection in Node.js What are good ways to prevent SQL injection? Found 'OR 1=1/* sql injection in my newsletter database How can prepared statements protect from SQL injection attacks? Why do we always prefer using parameters in SQL statements? SQL injection that gets around mysql_real_escape_string() Java - escape string to prevent SQL injection How does a PreparedStatement avoid or prevent SQL injection? How does the SQL injection from the "Bobby Tables" XKCD comic work? Are PDO prepared statements sufficient to prevent SQL injection?