For the past couple weeks I've been writing about how to protect your database from a SQL injection attack.  Today, we will keep the trend going by looking at how implicit unicode conversions can leave your data vulnerable.

Watch this week's video on YouTube

What's a homoglyph?

A homoglyph is a character that looks like another character. l  (lowercase "L") and 1  (the number) are considered homoglyphs.  So are O  (the letter) and 0  (the number).

Homoglpyhs can exist within a character set (like the Latin character set examples above) or they can exist between character sets.  For example, you may have the unicode apostrophe ʼ, which is a homoglyph to the Latin single quote character ' .

How does SQL Server handle unicode homoglyphs?

Funny you should ask.  If you pass in a unicode character to a non-unicode datatype (like char), SQL implicitly converts the unicode character to its closest resembling non-unicode homoglyph.

To see this in action, we can use the unicode apostrophe from the example above:

SELECT 
    CAST(N'ʼ' AS nchar) AS UnicodeChar, 
    CAST(N'ʼ' AS char) AS NonUnicodeChar

You can see in the second column SQL automatically converted the apostrophe to a single quote:

Although this implicit character conversion can be convenient for when you want to display unicode characters in a non-unicode character set, it can spell disaster for your SQL Server security.

Unicode Homoglyph SQL Injection

If you are already using sp_executesql or QUOTENAME() when building your dynamic SQL queries then you are safe from this type of SQL injection.

I know you aren't the kind of person who would ever write your own security functions when solid, safe, and tested functions like the above are available.  However, just this one time let's pretend you think you can outsmart a hacker by writing your own quote delimiting code.

Using the same dataset as last week, let's create a new stored procedure that is going to return some data from a user's profile:

DROP PROCEDURE IF EXISTS dbo.GetProfile
GO
CREATE PROCEDURE dbo.GetProfile
    @Username nvarchar(100)
AS
BEGIN
    -- Add quotes to escape injection...or not?
    SET @Username = REPLACE(@Username, '''','''''')

    DECLARE @Query varchar(max)

    SET @Query = 'SELECT 
                    FullName, 
                    JoinDate
                FROM
                    dbo.RegisteredUser
                WHERE
                    UserName = ''' + @Username + '''
                    '

    EXEC(@Query)
END
GO

Instead of using sp_executesql or QUOTENAME(), let's try to write our own clever REPLACE() function that will replace single quotes with two sets of single quotes.  This should, in theory, prevent SQL injection.

If we test out a "normal" attempt at SQL injection, you'll notice this logic works great.  Give yourself a pat on the back!

However, if we pass in a unicode apostrophe...:

The reason this happens is because we declared our @Query parameter as varchar instead of the unicode nvarchar.  When we build our dynamic SQL statement, SQL implicitly converts the nvarchar @Username parameter to the non-unicode varchar:

So if I replace apostrophes will that make me safe?

No.

I know it seems like black listing/replacing the unicode apostrophe would solve all of our problems.

And it would...in this scenario only.  There are more unicode homoglpyhs than just an apostrophe though.

Out of curiosity I wrote a script to search through the unicode character space to see what other homoglyphs exist:

DECLARE @FirstNumber INT=0;
-- number of possible characters in the unicode space
DECLARE @LastNumber INT=1114112;

WITH Numbers AS (
    SELECT @FirstNumber AS n
    UNION ALL
    SELECT n+1 FROM Numbers WHERE n+1<=@LastNumber
), UnicodeConversion AS (
SELECT
       n AS CharacterNumber,
       CASE CAST(NCHAR(n) as CHAR(1))
              WHEN '''' THEN NCHAR(n)
              WHEN ';' THEN NCHAR(n)
       END AS UnicodeCharacter,
       CAST(NCHAR(n) as CHAR(1)) AS ASCIICharacter
FROM Numbers
)
SELECT
       *
FROM
       UnicodeConversion
WHERE
       UnicodeCharacter IS NOT NULL
OPTION (MAXRECURSION 0)

Although the characters in the above screen shot might look similar, they are actually homoglyphs.

I decided to only search for single quotes and semi-colons since they are frequently used in SQL injection attacks, but this by no means is an extensive list of all of the characters you would want to blacklist.

Not only would it be very difficult to confidently blacklist every dangerous homoglyph, but new characters are being added to unicode all of the time so maintaining a black list would be a maintenance nightmare.  Especially if the person maintaining this code in the future isn't familiar with these types of injection attacks.

And don't be cheeky thinking you can filter out dangerous SQL keywords either - even if you REPLACE(@Username,'SELECT',''), just remember someone can come by and pass in a value like 'ŚεℒℇℂƮ'.

Conclusion

Don't write your own security functions - they will fail.

Your best protection against SQL injection is to not use dynamic SQL.  If you have to use dynamic SQL, then use either sp_executesql and QUOTENAME().

Protecting against SQL Injection Part 2

Watch this week's video on YouTube

Last week we talked about building dynamic SQL queries and how doing so might leave you open to SQL injection attacks.

In that post we examined how using sp_executesql to parameterize our dynamic SQL statements protects us.

Today, we are going to look at where sp_executesql falls short.

The problem with sp_executesql

Let's modify last week's final query to make our table name dynamic:

CREATE PROCEDURE dbo.sp_GetFullNameFromTable
        @ParmTableName varchar(100),
        @ParmUserName varchar(100)
AS
BEGIN
    DECLARE @FullQuery nvarchar(1000)
    SET @FullQuery = N'SELECT FullName FROM dbo.@TableName WHERE UserName = @UserName'

    DECLARE @ParmDefinition nvarchar(100) = N'@TableName varchar(100), @UserName varchar(100)';  

    EXEC sp_executesql @FullQuery, @ParmDefinition,  
                      @UserName = @ParmUserName,
                      @TableName = @ParmTableName;  

END

(there's an entirely different discussion to be had about whether you should allow table and column names to be dynamic, but we'll assume we need a dynamic table name for the rest of this post)

If we pass in a table name parameter value and execute this query, we'll be greeted with this error:

Yeah, sp_executesql doesn't like parameterizing a table names.

So how do we make table name dynamic and safe?

Unfortunately we have to fall back on SQL's EXEC command.

However, like we discussed last week, we need to be vigilant about what kind of user input we allow to be built as part of our query.

Assuming our app layer is already sanitizing as much of the user input as possible, here are some precautions we can take on the SQL side of the equation:

QUOTENAME()

If we wrap our user input parameters with QUOTENAME(), we'll achieve safety:

CREATE PROCEDURE dbo.sp_GetFullNameFromTableSanitized
    @ParmTableName varchar(100),
    @ParmUserName varchar(100)
AS
BEGIN
    DECLARE @FullQuery nvarchar(1000)
    SET @FullQuery = N'SELECT FullName FROM dbo.' + QUOTENAME(@ParmTableName) + ' WHERE UserName = @UserName'

    DECLARE @ParmDefinition nvarchar(100) = N'@UserName varchar(100)';  

    EXEC sp_executesql @FullQuery, @ParmDefinition,  
                      @UserName = @ParmUserName

END

This results in:

Although QUOTENAME() works well by adding brackets (by default) to sanitize quotes, it's downside is that it will only output strings up to 128 characters long. If you are expecting parameters with values longer than that, you will have to use something like REPLACE(@TableName,'''','''''') instead to delimit single quotes (however, rolling your own logic like this is really hard to do securely and not recommended).

EXECUTE AS

The account running any dynamic SQL queries should be locked down so that it won't be able to perform any operations you don't want it to do.

Taking this idea one step further, you can create another account with very limited permissions, and add EXECUTE AS to your stored procedure to run that code under the very limited account.

CREATE PROCEDURE dbo.sp_GetFullNameFromTableSanitized
    @ParmTableName varchar(100),
    @ParmUserName varchar(100)
    WITH EXECUTE AS 'LimitedUser'
AS
BEGIN
...

This won't prevent injection, but it will limit what the malicious user is able to do.

Watch this week's video on YouTube

Looking for a script to find possible SQL injection vulnerabilities on your server? Scroll to the bottom of this post.

OWASP names SQL injection as the #1 vulnerability for web applications. The infamy of this attack has even made its way into the popular XKCD comic.

What is SQL injection?

A SQL query is vulnerable to SQL injection if a user can run a query other than the one that was originally intended.

Sometimes SQL injection is easier to understand with an example. Let's use this table of registered users:

And then let's create a simple stored procedure that will query that table:

CREATE PROCEDURE dbo.sp_GetFullName
    @ParmUserName varchar(100)
AS
BEGIN
    DECLARE @FullQuery varchar(1000)
    SET @FullQuery = 'SELECT FullName FROM dbo.RegisteredUser WHERE UserName = ''' + @ParmUserName + ''''

    EXEC(@FullQuery);
END

The important thing to note in the query above is that we are generating a dynamic SQL statement; that is, we are building the SQL query string, and then we are executing it.

Imagine this stored procedure is running in order to display a "Welcome !" message in our app — a website visitor types in their@ParmUserName and we execute the stored procedure to return their full name.

Here's our code that calls the stored procedure:

EXEC dbo.sp_GetFullName 'TFly37'

And result:

Cool. No problems so far.

However, what if our user decides to pass in the following value as their username?

EXEC dbo.sp_GetFullName 'TFly37'' or 1=1 --'

This funny looking parameter value returns this:

AHHHHHH!!!!

This user just hacked our website and viewed all of the users in our table.

In this specific case only our user's full names were breached, but in other instances it's just as easy for more sensitive data like passwords, social security numbers, and bank account numbers to be revealed as well (If you are looking for some fun, search for "SQL injection" on the Have I been pwned? list of Pwned websites to see all of the companies that aren't doing a good job protecting your privacy).

So how did that example of SQL injection actually work?

Since our stored procedure executes a dynamically generated query, let's look at what that generated query actually looks like for both of the parameters that were passed in:

Even though TFly37'' or 1=1-- doesn't look like a intelligible input parameter, when its concatenated into our query it makes sense.

Our malicious user is basically writing their own SQL query, one that will return all of the names of our users instead of just their own. In many instances, the crafty hacker can go a step further and write additional injection code to reveal the contents of the entire user table (or other tables in the database)!

How do I prevent SQL injection?

Simple: don't concatenate unsanitized user input data to your SQL queries.

In this example, this is easy to do: simply rewrite the stored procedure to not use dynamic SQL:

CREATE PROCEDURE dbo.sp_GetFullNameSafe
    @ParmUserName varchar(100)
AS
BEGIN
    SELECT FullName FROM dbo.RegisteredUser WHERE UserName =  @ParmUserName
END

When you don't have dynamic SQL, you can't be affected by SQL injection attempts.

Avoiding dynamic SQL isn't always possible though, so what can we do in those cases? Use sp_executesql:

CREATE PROCEDURE dbo.sp_GetFullNameSafe2
    @ParmUserName varchar(100)
AS
BEGIN
    DECLARE @FullQuery nvarchar(1000)
    SET @FullQuery = N'SELECT FullName FROM dbo.RegisteredUser WHERE UserName = @UserName'

    DECLARE @ParmDefinition nvarchar(100) = N'@UserName varchar(100)';  

    EXEC sp_executesql @FullQuery, @ParmDefinition,  
                      @UserName = @ParmUserName;  
END

sp_executesql allows for parameterization of your dynamic SQL, removing the possibility of unwanted code injection.

Are any of my current SQL queries vulnerable to SQL injection attacks?

So SQL injection is really bad and you don't want to become like Sony or Yahoo. How do we check to see if we have any vulnerable queries on our server?

I wrote the query below to help you start searching. It is not perfect, but it does act as a good starting point for auditing potentially injectable queries:

-- This file tries to find stored procedures and functions that *may* be vulnerable to SQL injection attacks.

-- It works by searching your database for occurences of "+" signs followed by "@", indicating that SQL parameters
-- might be getting concatenated to a dynamic SQL string.  It also checks for the existence of 'EXEC' to see if any
-- strings are being executed.

-- Not every result returned will be susceptible to SQL injection, however they should all be examined to see if they are vulnerable.

-- Originally fromn: https://github.com/bertwagner/SQLServer/blob/master/SQL%20Injection%20Vulnerabilities.sql

SET TRANSACTION ISOLATION LEVEL READ UNCOMMITTED

SELECT
    ROUTINE_CATALOG,
    ROUTINE_SCHEMA,
    ROUTINE_NAME,
    ROUTINE_TYPE,
    ROUTINE_DEFINITION
FROM
    INFORMATION_SCHEMA.ROUTINES
WHERE
    REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(ROUTINE_DEFINITION,CHAR(0),''),CHAR(9),''),CHAR(10),''),CHAR(11),''),CHAR(12),''),CHAR(13),''),CHAR(14),''),CHAR(160),''),' ','')
        LIKE '%+@%'
    AND 
    ( -- Only if executes a dynamic string
        ROUTINE_DEFINITION LIKE '%EXEC(%'
        OR ROUTINE_DEFINITION LIKE '%EXECUTE%'
        OR ROUTINE_DEFINITION LIKE '%sp_executesql%'
    )

I've found this script useful for myself, however if you find any issues with it please let me know, thanks!