sse - Dave's Blog

There are two main differences in terms of security between a JavaScript UWP app and the Edge browser:

Process Model

A JavaScript UWP app has one process (technically not true with background tasks and other edge cases but ignoring that for the moment) that runs in the corresponding appcontainer defined by the app's appx manifest. This one process is where edgehtml is loaded and is rendering HTML, talking to the network, and executing script. Specifically, the UWP main UI thread is the one where your script is running and calling into WinRT.

In the Edge browser there is a browser process running in the same appcontainer defined by its appx manifest, but there are also tab processes. These tab processes are running in restricted app containers that have fewer appx capabilities. The browser process has XAML loaded and coordinates between tabs and handles some (non-WinRT) brokering from the tab processes. The tab processes load edgehtml and that is where they render HTML, talk to the network and execute script.

There is no way to configure the JavaScript UWP app's process model but using WebViews you can approximate it. You can create out of process WebViews and to some extent configure their capabilities, although not to the same extent as the browser. The WebView processes in this case are similar to the browser's tab processes. See the MSWebViewProcess object for configuring out of process WebView creation. I also implemented out of proc WebView tabs in my JSBrowser fork.

ApplicationContentUriRules

The ApplicationContentUriRules (ACUR) section of the appx manifest lets an application define what URIs are considered app code. See a previous post for the list of ACUR effects.

Notably app code is able to access WinRT APIs. Because of this, DOM security restrictions are loosended to match what is possible with WinRT.

Privileged DOM APIs like geolocation, camera, mic etc require a user prompt in the browser before use. App code does not show the same browser prompt. There still may be an OS prompt – the same prompt that applies to any UWP app, but that’s usually per app not per origin.

App code also gets to use XMLHttpRequest or fetch to access cross origin content. Because UWP apps have separate state, cross origin here might not mean much to an attacker unless your app also has the user login to Facebook or some other interesting cross origin target.

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.

[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.

JSBrowser is a basic browser built as a Win10 JavaScript UWP app around the WebView HTML element. Its fun and relatively simple to implement tiny browser features in JavaScript and in this post I'm implementing crash resistance.

The normal DOM mechanisms for creating an HTML WebView create an in-process WebView, in which the WebView runs on a unique UI thread. But we can use the MSWebView constructor instead to create an out-of-process WebView in which the WebView runs in its own distinct WebView process. Unlike an in-process WebView, Web content running in an out-of-process WebView can only crash the WebView process and not the app process.

        this.replaceWebView = () => {
            let webview = document.querySelector("#WebView");
            // Cannot access webview.src - anything that would need to communicate with the webview process may fail
            let oldSrc = browser.currentUrl;
            const webviewParent = webview.parentElement;
            webviewParent.removeChild(webview);
            webview = new MSWebView();
            Object.assign(this, {
                "webview": webview
            });
            webview.setAttribute("id", "WebView");

            // During startup our currentUrl field is blank. If the WebView has crashed 
            // and we were on a URI then we may obtain it from this property.
            if (browser.currentUrl && browser.currentUrl != "") {
                this.trigger("newWebview");
                this.navigateTo(browser.currentUrl);
            }
            webviewParent.appendChild(webview);

I run replaceWebView during startup to replace the in-process WebView created via HTML markup with an out-of-process WebView. I could be doing more to dynamically copy styles, attributes, etc but I know what I need to set on the WebView and just do that.

When a WebView process crashes the corresponding WebView object is no longer useful and a new WebView element must be created. In fact if the old WebView object is used it may throw and will no longer have valid state. Accordingly when the WebView crashes I run replaceWebView again. Additionally, I need to store the last URI we've navigated to (browser.currentUrl in the above) since the crashed WebView object won't know what URI it is on after it crashes.

            webview.addEventListener("MSWebViewProcessExited", () => { 
                if (browser.currentUrl === browser.lastCrashUrl) {                   ++browser.lastCrashUrlCrashCount; 
                } 
                else { 
                    browser.lastCrashUrl = browser.currentUrl; 
                    browser.lastCrashUrlCrashCount = 1; 
                } 
                // If we crash again and again on the same URI, maybe stop trying to load that URI. 
                if (browser.lastCrashUrlCrashCount >= 3) { 
                    browser.lastCrashUrl = ""; 
                    browser.lastCrashUrlCrashCount = 0; 
                    browser.currentUrl = browser.startPage; 
                } 
                this.replaceWebView(); 
            }); 

I also keep track of the last URI that we recovered and how many times we've recovered that same URI. If the same URI crashes more than 3 times in a row then I assume that it will keep happening and I navigate to the start URI instead.

Application Content URI Rules (ACUR from now on) defines the bounds on the web that make up a Microsoft Store application. The previous blog post discussed the syntax of the Rule's Match attribute and this time I'll write about the interactions between the Rules elements.

Order

A single ApplicationContentUriRules element may have up to 100 Rule child elements. When determining if a navigation URI matches any of the ACUR the last Rule in the list with a matching match wildcard URI is used. If that Rule is an include rule then the navigation URI is determined to be an application content URI and if that Rule is an exclude rule then the navigation rule is not an application content URI. For example:

Rule Type='include' Match='https://example.com/'/
Rule Type='exclude' Match='https://example.com/'/

Given the above two rules in that order, the navigation URI https://example.com/ is not an application content URI because the last matching rule is the exclude rule. Reverse the order of the rules and get the opposite result.

WindowsRuntimeAccess

In addition to determining if a navigation URI is application content or not, a Rule may also confer varying levels of WinRT access via the optional WindowsRuntimeAccess attribute which may be set to 'none', 'allowForWeb', or 'all'. If a navigation URI matches multiple different include rules only the last rule is applied even as it applies to the WindowsRuntimeAccess attribute. For example:

Rule Type='include' Match='https://example.com/' WindowsRuntimeAccess='none'/
Rule Type='include' Match='https://example.com/' WindowsRuntimeAccess='all'/

Given the above two rules in that order, the navigation URI https://example.com/ will have access to all WinRT APIs because the last matching rule wins. Reverse the rule order and the navigation URI https://example.com/ will have no access to WinRT. There is no summation or combining of multiple matching rules - only the last matching rule wins.

Parsing WinMD files, the containers of WinRT API metadata, is relatively simple using the appropriate .NET reflection APIs. However, figuring out which reflection APIs to use is not obvious. I've got a completed C sharp class parsing WinMD files that you can check out for reference.

Use System.Reflection.Assembly.ReflectionOnlyLoad to load the WinMD file. Don't use the normal load methods because the WinMD files contain only metadata. This will load up info about APIs defined in that WinMD, but any references to types outside of that WinMD including types found in the normal OS system WinMD files must be resolved by the app code via the System.Reflection.InteropServices.WindowsRuntimeMetadata.ReflectionOnlyNamespaceResolve event.

In this event handler you must resolve the unknown namespace reference by adding an assembly to the NamespaceResolveEventArgs's ResolvedAssemblies property. If you're only interested in OS system WinMD files you can use System.Reflection.InteropServices.WindowsRuntimeMetadata.ResolveNamespace to turn a namespace into the expected OS system WinMD path and turn that path into an assembly with ReflectionOnlyLoad.

The other day I had to debug a JavaScript UWA that was failing when trying to use an undefined property. In a previous OS build this code would run and the property was defined. I wanted something similar to windbg/cdb's ba command that lets me set a breakpoint on read or writes to a memory location so I could see what was creating the object in the previous OS build and what that code was doing now in the current OS build. I couldn't find such a breakpoint mechanism in Visual Studio or F12 so I wrote a little script to approximate JavaScript data breakpoints.

The script creates a stub object with a getter and setter. It actually performs the get or set but also calls debugger; to break in the debugger. In order to handle my case of needing to break when window.object1.object2 was created or accessed, I further had it recursively set up such stub objects for the matching property names.

Its not perfect because it is an enumerable property and shows up in hasOwnProperty and likely other places. But for your average code that checks for the existence of a property via if (object.property) it works well.

TL;DR: Web content in a JavaScript Windows Store app or WebView in a Windows Store app that has full access to WinRT also gets to use XHR unrestricted by cross origin checks.

By default web content in a WebView control in a Windows Store App has the same sort of limitations as that web content in a web browser. However, if you give the URI of that web content full access to WinRT, then the web content also gains the ability to use XMLHttpRequest unrestricted by cross origin checks. This means no CORS checks and no OPTIONS requests. This only works if the web content's URI matches a Rule in the ApplicationContentUriRules of your app's manifest and that Rule declares WindowsRuntimeAccess="all". If it declares WinRT access as 'None' or 'AllowForWebOnly' then XHR acts as it normally does.

In terms of security, if you've already given a page access to all of WinRT which includes the HttpRequest class and other networking classes that don't perform cross origin checks, then allowing XHR to skip CORS doesn't make things worse.

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'.

I've made a Unicode Clock in JavaScript.

Unicode has code points for all 30 minute increments of clock faces. This is a simple project to display the one closest to the current time written in JavaScript.

Because the code points are all above 0xFFFF, I make use of some ES6 additions. I use the \u{XXXXXX} style escape sequence since the old style JavaScript escape sequence \uXXXX only supports code points up to 0xFFFF. I also use the method String.codePointAt rather than String.charCodeAt because the code points larger than 0xFFFF are represented in JavaScript strings using surrogate pairs and charCodeAt gives the surrogate value rather than codePointAt which gives the code point represented by the pair of surrogates.

"🕛".codePointAt(0)
128347
"🕛".charCodeAt(0)
55357

🕐🕑🕒🕓🕔🕕🕖🕗🕘🕙🕚🕛🕜🕝🕞🕟🕠🕡🕢🕣🕤🕥🕦🕧

The ordering of the code points does not make it simple to do this. I initially guessed the first code point in the range would be 12:00 followed by 12:30, 1:00 and so on. But actually 1:00 is first followed by all the on the hour times then all the half hour times.

MSDN covers the topic of JavaScript and WinRT type conversions provided by Chakra (JavaScript Representation of Windows Runtime Types and Considerations when Using the Windows Runtime API), but for the questions I get about it I’ll try to lay out some specifics of that discussion more plainly. I’ve made a TL;DR JavaScript types and WinRT types summary table.

Chakra, the JavaScript engine powering the Edge browser and JavaScript Windows Store apps, does the work to project WinRT into JavaScript. It is responsible for, among other things, converting back and forth between JavaScript types and WinRT types. Some basics are intuitive, like a JavaScript string is converted back and forth with WinRT’s string representation. For other basic types check out the MSDN links at the top of the page. For structs, interface instances, class instances, and objects things are more complicated.

A struct, class instance, or interface instance in WinRT is projected into JavaScript as a JavaScript object with corresponding property names and values. This JavaScript object representation of a WinRT type can be passed into other WinRT APIs that take the same underlying type as a parameter. This JavaScript object is special in that Chakra keeps a reference to the underlying WinRT object and so it can be reused with other WinRT APIs.

However, if you start with plain JavaScript objects and want to interact with WinRT APIs that take non-basic WinRT types, your options are less plentiful. You can use a plain JavaScript object as a WinRT struct, so long as the property names on the JavaScript object match the WinRT struct’s. Chakra will implicitly create an instance of the WinRT struct for you when you call a WinRT method that takes that WinRT struct as a parameter and fill in the WinRT struct’s values with the values from the corresponding properties on your JavaScript object.

// C# WinRT component
        public struct ExampleStruct
        {
            public string String;
            public int Int;
        }

        public sealed class ExampleStructContainer
        {
            ExampleStruct value;
            public void Set(ExampleStruct value)
            {
                this.value = value;
            }

            public ExampleStruct Get()
            {
                return this.value;
            }
        }

// JS code
        var structContainer = new ExampleWinRTComponent.ExampleNamespace.ExampleStructContainer();
        structContainer.set({ string: "abc", int: 123 });
        console.log("structContainer.get(): " + JSON.stringify(structContainer.get()));
        // structContainer.get(): {"string":"abc","int":123}

You cannot have a plain JavaScript object and use it as a WinRT class instance or WinRT interface instance. Chakra does not provide such a conversion even with ES6 classes.

You cannot take a JavaScript object with arbitrary property names that are unknown at compile time and don’t correspond to a specific WinRT struct and pass that into a WinRT method. If you need to do this, you have to write additional JavaScript code to explicitly convert your arbitrary JavaScript object into an array of property name and value pairs or something else that could be represented in WinRT.

However, the other direction you can do. An instance of a Windows.Foundation.Collections.IPropertySet implementation in WinRT is projected into JavaScript as a JavaScript object with property names and values corresponding to the key and value pairs in the IPropertySet. In this way you can project a WinRT object as a JavaScript object with arbitrary property names and types. But again, the reverse is not possible. Chakra will not convert an arbitrary JavaScript object into an IPropertySet.

// C# WinRT component
        public sealed class PropertySetContainer
        {
            private Windows.Foundation.Collections.IPropertySet otherValue = null;

            public Windows.Foundation.Collections.IPropertySet other
            {
                get
                {
                    return otherValue;
                }
                set
                {
                    otherValue = value;
                }
            }
        }

        public sealed class PropertySet : Windows.Foundation.Collections.IPropertySet
        {
            private IDictionary map = new Dictionary();

            public PropertySet()
            {
                map.Add("abc", "def");
                map.Add("ghi", "jkl");
                map.Add("mno", "pqr");
            }
            // ... rest of PropertySet implementation is simple wrapper around the map member.
            

// JS code
    var propertySet = new ExampleWinRTComponent.ExampleNamespace.PropertySet();
    console.log("propertySet: " + JSON.stringify(propertySet));
    // propertySet: {"abc":"def","ghi":"jkl","mno":"pqr"}

    var propertySetContainer = new ExampleWinRTComponent.ExampleNamespace.PropertySetContainer();
    propertySetContainer.other = propertySet;
    console.log("propertySetContainer.other: " + JSON.stringify(propertySetContainer.other));
    // propertySetContainer.other: {"abc":"def","ghi":"jkl","mno":"pqr"}

    try {
        propertySetContainer.other = { "123": "456", "789": "012" };
    }
    catch (e) {
        console.error("Error setting propertySetContainer.other: " + e);
        // Error setting propertySetContainer.other: TypeError: Type mismatch
}

There’s also no way to implicitly convert a DOM object into a WinRT type. If you want to write third party WinRT code that interacts with the DOM, you must do so indirectly and explicitly in JavaScript code that is interacting with your third party WinRT. You’ll have to extract the information you want from your DOM objects to pass into WinRT methods and similarly have to pass messages out from WinRT that say what actions the JavaScript should perform on the DOM.