server - Dave's Blog

If you're developing with the new Windows.Web.UI.Interop.WebViewControl you may have noticed you cannot navigate to localhost HTTP servers. This is because the WebViewControl's WebView process is a UWP process. All UWP processes by default cannot use the loopback adapter as a security precaution. For development purposes you can allow localhost access using the checknetisolation command line tool on the WebViewControl's package just as you can for any other UWP app. The command should be the following:

checknetisolation loopbackexempt -a -n=Microsoft.Win32WebViewHost_cw5n1h2txyewy

As a warning checknetisolation is not good on errors. If you attempt to add a package but get its package family name wrong, checknetisolation just says OK:

C:\Users\davris>checknetisolation LoopbackExempt -a -n=Microsoft.BingWeather_4.21.2492.0_x86__8wekyb3d8bbwe
OK.

And if you then list the result of the add with the bad name you'll see the following:

[1] -----------------------------------------------------------------
    Name: AppContainer NOT FOUND
    SID:  S-1-15-...

There's also a UI tool for modifying loopback exemption for packages available on GitHub and also one available with Fiddler.

As an additional note, I mentioned above you can try this for development. Do not do this in shipping products as this turns off the security protection for any consumer of the WebViewControl.

I've put my WPAD Fiddler extension source and the installer on GitHub.

Six years ago I made a WPAD DHCP server Fiddler extension (described previously and previously). The extension runs a WPAD DHCP server telling any clients that connect to connect to the running Fiddler instance. I've finally got around to putting the source on GitHub. I haven't touched it in five or so years so this is either for posterity or education or something.

2016-Nov-5: Updated post on using Let's Encrypt with NearlyFreeSpeech.net

I use NearlyFreeSpeech.net for my webhosting for my personal website and I've just finished setting up TLS via Let's Encrypt. The process was slightly more complicated than what you'd like from Let's Encrypt. So for those interested in doing the same on NearlyFreeSpeech.net, I've taken the following notes.

The standard Let's Encrypt client requires su/sudo access which is not available on NearlyFreeSpeech.net's servers. Additionally NFSN's webserver doesn't have any Let's Encrypt plugins installed. So I used the Let's Encrypt Without Sudo client. I followed the instructions listed on the tool's page with the addition of providing the "--file-based" parameter to sign_csr.py.

One thing the script doesn't produce is the chain file. But this topic "Let's Encrypt - Quick HOWTO for NSFN" covers how to obtain that:

curl -o domain.chn https://letsencrypt.org/certs/lets-encrypt-x1-cross-signed.pem

Now that you have all the required files, on your NFSN server make the directory /home/protected/ssl and copy your files into it. This is described in the NFSN topic provide certificates to NFSN. After copying the files and setting their permissions as described in the previous link you submit an assistance request. For me it was only 15 minutes later that everything was setup.

After enabling HTTPS I wanted to have all HTTP requests redirect to HTTPS. The normal Apache documentation on how to do this doesn't work on NFSN servers. Instead the NFSN FAQ describes it in "redirect http to https and HSTS". You use the X-Forwarded-Proto instead of the HTTPS variable because of how NFSN's virtual hosting is setup.

RewriteEngine on
RewriteCond %{HTTP:X-Forwarded-Proto} !https
RewriteRule ^.*$ https://%{SERVER_NAME}%{REQUEST_URI} [L,R=301]

Turning on HSTS is as simple as adding the HSTS HTTP header. However, the description in the above link didn't work because my site's NFSN realm isn't on the latest Apache yet. Instead I added the following to my .htaccess. After I'm comfortable with everything working well for a few days I'll start turning up the max-age to the recommended minimum value of 180 days.

Header set Strict-Transport-Security "max-age=3600;" 

Finally, to turn on CSP I started up Fiddler with my CSP Fiddler extension. It allows me to determine the most restrictive CSP rules I could apply and still have all resources on my page load. From there I found and removed inline script and some content loaded via http and otherwise continued tweaking my site and CSP rules.

After I was done I checked out my site on SSL Lab's SSL Test to see what I might have done wrong or needed improving. The first time I went through these steps I hadn't included the chain file which the SSL Test told me about. I was able to add that file to the same files I had already previously generated from the Let's Encrypt client and do another NFSN assistance request and 15 minutes later the SSL Test had upgraded me from 'B' to 'A'.

Level 8 of the Stripe CTF is a password server that returns success: true if and only if the password provided matches the password stored directly via a RESTful API and optionally indirectly via a callback URI. The solution is side channel attack like a timing attack but with ports instead of time.

(I found this in my drafts folder and had intended to post a while ago.)

Code

    def nextServerCallback(self, data):
        parsed_data = json.loads(data)
        # Chunk was wrong!
        if not parsed_data['success']:
            # Defend against timing attacks
            remaining_time = self.expectedRemainingTime()
            self.log_info('Going to wait %s seconds before responding' %
                          remaining_time)
            reactor.callLater(remaining_time, self.sendResult, False)
            return

        self.checkNext()

Issue

The password server breaks the target password into four pieces and stores each on a different server. When a password request is sent to the main server it makes requests to the sub-servers for each part of the password request. It does this in series and if any part fails, then it stops midway through. Password requests may also be made with corresponding URI callbacks and after the server decides on the password makes an HTTP request on the provided URI callbacks saying if the password was success: true or false.
A timing attack looks at how long it took for a password to be rejected and longer times could mean a longer prefix of the password was correct allowing for a directed brute force attack. Timing attacks are prevented in this case by code on the password server that attempts to wait the same amount of time, even if the first sub-server responds with false. However, the server uses sequential outgoing port numbers shared between the requests to the sub-servers and the callback URIs. Accordingly, we can examine the port numbers on our callback URIs to direct a brute force attack.
If the password provided is totally incorrect then the password server will contact one sub-server and then your callback URI. So if you see the remote server's port number go up by two when requesting your callback URI, you know the password is totally incorrect. If by three then you know the first fourth of the password is correct and the rest is incorrect. If by four then two fourths of the password is correct. If by five then four sub-servers were contacted so you need to rely on the actual content of the callback URI request of 'success: true' or 'false' since you can't tell from the port change if the password was totally correct or not.
The trick in the real world is false positives. The port numbers are sequential over the system, so if the password server is the only thing making outgoing requests then its port numbers will also be sequential, however other things on the system can interrupt this. This means that the password server could contact three sub-servers and normally you'd see the port number increase by four, but really it could increase by four or more because of other things running on the system. To counteract this I ran in cycles: brute forcing the first fourth of the password and removing any entry that gets a two port increase and keeping all others. Eventually I could remove all but the correct first fourth of the password. And so on for the next parts of the password.
I wrote my app to brute force this in Python. This was my first time writing Python code so it is not pretty.

Level 7 of the Stripe CTF involved running a length extension attack on the level 7 server's custom crypto code.

Code

@app.route('/logs/')
@require_authentication
def logs(id): 
    rows = get_logs(id) 
    return render_template('logs.html', logs=rows) 

...

def verify_signature(user_id, sig, raw_params):
    # get secret token for user_id
    try:
        row = g.db.select_one('users', {'id': user_id})
    except db.NotFound:
        raise BadSignature('no such user_id')
    secret = str(row['secret'])

    h = hashlib.sha1()
    h.update(secret + raw_params)
    print 'computed signature', h.hexdigest(), 'for body', repr(raw_params)
    if h.hexdigest() != sig:
        raise BadSignature('signature does not match')
    return True

Issue

The level 7 web app is a web API in which clients submit signed RESTful requests and some actions are restricted to particular clients. The goal is to view the response to one of the restricted actions. The first issue is that there is a logs path to display the previous requests for a user and although the logs path requires the client to be authenticatd, it doesn't restrict the logs you view to be for the user for which you are authenticated. So you can manually change the number in the '/logs/[#]' to '/logs/1' to view the logs for the user ID 1 who can make restricted requests. The level 7 web app can be exploited with replay attacks but you won't find in the logs any of the restricted requests we need to run for our goal. And we can't just modify the requests because they are signed.

However they are signed using their own custom signing code which can be exploited by a length extension attack. All Merkle–Damgård hash algorithms (which includes MD5, and SHA) have the property that if you hash data of the form (secret + data) where data is known and the length but not content of secret is known you can construct the hash for a new message (secret + data + padding + newdata) where newdata is whatever you like and padding is determined using newdata, data, and the length of secret. You can find a sha-padding.py script on VNSecurity blog that will tell you the new hash and padding per the above. With that I produced my new restricted request based on another user's previous request. The original request was the following.

count=10&lat=37.351&user_id=1&long=%2D119.827&waffle=eggo|sig:8dbd9dfa60ef3964b1ee0785a68760af8658048c

The new request with padding and my new content was the following.

count=10&lat=37.351&user_id=1&long=%2D119.827&waffle=eggo%80%02%28&waffle=liege|sig:8dbd9dfa60ef3964b1ee0785a68760af8658048c

My new data in the new request is able to overwrite the waffle parameter because their parser fills in a map without checking if the parameter existed previously.

Notes

Code review red flags included custom crypto looking code. However I am not a crypto expert and it was difficult for me to find the solution to this level.

Level 5 of the Stripe CTF revolved around a design issue in an OpenID like protocol.

Code

    def authenticated?(body)
      body =~ /[^\w]AUTHENTICATED[^\w]*$/
    end

...

      if authenticated?(body)
        session[:auth_user] = username
        session[:auth_host] = host
        return "Remote server responded with: #{body}." \
               " Authenticated as #{username}@#{host}!"

Issue

This level is an implementation of a federated identity protocol. You give it an endpoint URI and a username and password, it posts the username and password to the endpoint URI, and if the response is 'AUTHENTICATED' then access is allowed. It is easy to be authenticated on a server you control, but this level requires you to authenticate from the server running the level. This level only talks to stripe CTF servers so the first step is to upload a document to the level 2 server containing the text 'AUTHENTICATED' and we can now authenticate on a level 2 server. Notice that the level 5 server will dump out the content of the endpoint URI and that the regexp it uses to detect the text 'AUTHENTICATED' can match on that dump. Accordingly I uploaded an authenticated file to

https://level02-2.stripe-ctf.com/user-ajvivlehdt/uploads/authenticated

Using that as my endpoint URI means authenticating as level 2. I can then choose the following endpoint URI to authenticate as level 5.

https://level05-1.stripe-ctf.com/user-qtoyekwrod/?pingback=https%3A%2F%2Flevel02-2.stripe-ctf.com%2Fuser-ajvivlehdt%2Fuploads%2Fauthenticated&username=a&password=a

Navigating to that URI results in the level 5 server telling me I'm authenticated as level 2 and lists the text of the level 2 file 'AUTHENTICATED'. Feeding this back into the level 5 server as my endpoint URI means level 5 seeing 'AUTHENTICATED' coming back from a level 5 URI.

Notes

I didn't see any particular code review red flags, really the issue here is that the regular expression testing for 'AUTHENTICATED' is too permisive and the protocol itself doesn't do enough. The protocol requires only a set piece of common literal text to be returned which makes it easy for a server to accidentally fall into authenticating. Having the endpoint URI have to return variable text based on the input would make it much harder for a server to accidentally authenticate.

Level 4 and level 6 of the Stripe CTF had solutions around XSS.

Level 4

Code

> Registered Users 

<%@registered_users.each do |user| %>
<%last_active = user[:last_active].strftime('%H:%M:%S UTC') %>
<%if @trusts_me.include?(user[:username]) %>

• 
<%= user[:username] %>
(password: <%= user[:password] %>, last active <%= last_active %>)


Issue

The level 4 web application lets you transfer karma to another user and in doing so you are also forced to expose your password to that user. The main user page displays a list of users who have transfered karma to you along with their password. The password is not HTML encoded so we can inject HTML into that user's browser. For instance, we could create an account with the following HTML as the password which will result in XSS with that HTML:

This HTML runs script that uses jQuery to post to the transfer URI resulting in a transfer of karma from the attacked user to the attacker user, and also the attacked user's password.

Notes

Code review red flags in this case included lack of encoding when using user controlled content to create HTML content, storing passwords in plain text in the database, and displaying passwords generally. By design the web app shows users passwords which is a very bad idea.

Level 6

Code

...

    def self.safe_insert(table, key_values)
      key_values.each do |key, value|
        # Just in case people try to exfiltrate
        # level07-password-holder's password
        if value.kind_of?(String) &&
            (value.include?('"') || value.include?("'"))
          raise "Value has unsafe characters"
        end
      end

      conn[table].insert(key_values)
    end

Issue

This web app does a much better job than the level 4 app with HTML injection. They use encoding whenever creating HTML using user controlled data, however they don't use encoding when injecting JSON data into script (see post_data initialization above). This JSON data is the last five most recent messages sent on the app so we get to inject script directly. However, the system also ensures that no strings we write contains single or double quotes so we can't get out of the string in the JSON data directly. As it turns out, HTML lets you jump out of a script block using no matter where you are in script. For instance, in the middle of a value in some JSON data we can jump out of script. But we still want to run script, so we can jump right back in. So the frame so far for the message we're going to post is the following:

I was the 546th person to complete Stripe's web security CTF and again had a ton of fun applying my theoretical knowledge of web security issues to the (semi-)real world. As I went through the levels I thought about what red flags jumped out at me (or should have) that I could apply to future code reviews:

Level	Issue	Code Review Red Flags
0	Simple SQL injection	No encoding when constructing SQL command strings. Constructing SQL command strings instead of SQL API
1	extract($_GET);	No input validation.
2	Arbitrary PHP execution	No input validation. Allow file uploads. File permissions modification.
3	Advanced SQL injection	Constructing SQL command strings instead of SQL API.
4	HTML injection, XSS and CSRF	No encoding when constructing HTML. No CSRF counter measures. Passwords stored in plain text. Password displayed on site.
5	Pingback server doesn't need to opt-in	n/a - By design protocol issue.
6	Script injection and XSS	No encoding while constructing script. Deny list (of dangerous characters). Passwords stored in plain text. Password displayed on site.
7	Length extension attack	Custom crypto code. Constructing SQL command string instead of SQL API.
8	Side channel attack	Password handling code. Timing attack mitigation too clever.

More about each level in the future.

Stripe is running a web security capture the flag - a series of increasingly difficult web security exploit challenges. I've finished it and had a lot of fun. Working on a web browser I knew the theory of these various web based attacks, but this was my first chance to put theory into practice with:

• No adverse consequences
• Knowledge that there is a fun security exploit to find
• Access to the server side source code

Here's a blog post on the CTF behind the scenes setup which has many impressive features including phantom users that can be XSS/CSRF'ed.

I'll have another post on my difficulties and answers for the CTF levels after the contest is over on Wed, but if you're looking for hints, try out the CTF chatroom or the level specific CTF chatroom.

There's a lot of name reuse in HTTP compression so I've made the following to help myself keep it straight.

HTTP Content Coding Token	gzip	deflate	compress
	An encoding format produced by the file compression program "gzip" (GNU zip)	The "zlib" format as described in RFC 1950.	The encoding format produced by the common UNIX file compression program "compress".
Data Format	GZIP file format	ZLIB Compressed Data Format	The compress program's file format
Compression Method	Deflate compression method		LZW
	Deflate consists of LZ77 and Huffman coding

Compress doesn't seem to be supported by popular current browsers, possibly due to its past with patents.

Deflate isn't done correctly all the time. Some servers would send the deflate data format instead of the zlib data format and at least some versions of Internet Explorer expect deflate data format instead of zlib data format.

One of the more limiting issues of writing client side script in the browser is the same origin limitations of XMLHttpRequest. The latest version of all browsers support a subset of CORS to allow servers to opt-in particular resources for cross-domain access. Since IE8 there's XDomainRequest and in all other browsers (including IE10) there's XHR L2's cross-origin request features. But the vast majority of resources out on the web do not opt-in using CORS headers and so client side only web apps like a podcast player or a feed reader aren't doable.

One hack-y way around this I've found is to use YQL as a CORS proxy. YQL applies the CORS header to all its responses and among its features it allows a caller to request an arbitrary XML, HTML, or JSON resource. So my network helper script first attempts to access a URI directly using XDomainRequest if that exists and XMLHttpRequest otherwise. If that fails it then tries to use XDR or XHR to access the URI via YQL. I wrap my URIs in the following manner, where type is either "html", "xml", or "json":

        yqlRequest = function(uri, method, type, onComplete, onError) {
            var yqlUri = "http://query.yahooapis.com/v1/public/yql?q=" + 
                encodeURIComponent("SELECT * FROM " + type + ' where url="' + encodeURIComponent(uri) + '"');

            if (type == "html") {
                yqlUri += encodeURIComponent(" and xpath='/*'");
            }
            else if (type == "json") {
                yqlUri += "&callback=&format=json";
            }
            ...

This also means I can get JSON data itself without having to go through JSONP.

Shortly after joining the Internet Explorer team I got a bug from a PM on a popular Microsoft web server product that I'll leave unnamed (from now on UWS). The bug said that IE was handling empty path segments incorrectly by not removing them before resolving dotted path segments. For example UWS would do the following:

A.1. http://example.com/a/b//../
A.2. http://example.com/a/b/../
A.3. http://example.com/a/

In step 1 they are given a URI with dotted path segment and an empty path segment. In step 2 they remove the empty path segment, and in step 3 they resolve the dotted path segment. Whereas, given the same initial URI, IE would do the following:

B.1. http://example.com/a/b//../
B.2. http://example.com/a/b/

IE simply resolves the dotted path segment against the empty path segment and removes them both. So, how did I resolve this bug? As "By Design" of course!

The URI RFC allows path segments of zero length and does not assign them any special meaning. So generic user agents that intend to work on the web must not treat an empty path segment any different from a path segment with some text in it. In the case above IE is doing the correct thing.

That's the case for generic user agents, however servers may decide that a URI with an empty path segment returns the same resource as a the same URI without that empty path segment. Essentially they can decide to ignore empty path segments. Both IIS and Apache work this way and thus return the same resource for the following URIs:

http://exmaple.com/foo//bar///baz
http://example.com/foo/bar/baz

The issue for UWS is that it removes empty path segments before resolving dotted path segments. It must follow normal URI procedure before applying its own additional rules for empty path segments. Not doing that means they end up violating URI equivalency rules: URIs (A.1) and (B.2) are equivalent but UWS will not return the same resource for them.

As it turns out the WPAD Server Fiddler Extension I made a while back actually has a non-malicious purpose. Apparently its useful for debugging HTTP on the WP7 phone (or so I'm told). Anyway I took some requests and I've fixed a few minor bugs (start button not updating correctly), changed the dialog to be a Fiddler tab so you can use it non-modally, and the WPAD server is now always off when Fiddler starts.