browsers, perf, privacy, security

Private Mode Browsers

InPrivate Mode was introduced in Internet Explorer 8 with the goal of helping users improve their privacy against both local and remote threats. Safari introduced a privacy mode in 2005.

All leading browsers offer a “Private Mode” and they all behave in the same general ways.

HTTP Caching

While in Private mode, browsers typically ignore any previously cached resources and cookies. Similarly, the Private mode browser does not preserve any cached resources beyond the end of the browser session. These features help prevent a revisited website from trivially identifying a returning user (e.g. if the user’s identity were cached in a cookie or JSON file on the client) and help prevent “traces” that might be seen by a later user of the device.

In Firefox’s and Chrome’s Private modes, a memory-backed cache container is used for the HTTP cache, and its memory is simply freed when the browser session ends. Unfortunately, WinINET never implemented a memory cache, so in Internet Explorer InPrivate sessions, data is cached in a special WinINET cache partition on disk which is “cleaned up” when the InPrivate session ends.

Because this cleanup process may be unreliable, in 2017, Edge made a change to simply disable the cache while running InPrivate, a design decision with significant impact on the browser’s network utilization and performance. For instance, consider the scenario of loading an image gallery that shows one large picture per page and clicking “Next” ten times:

InPrivateVsRegular

Because the gallery reuses some CSS, JavaScript, and images across pages, disabling the HTTP cache means that these resources must be re-downloaded on every navigation, resulting in 50 additional requests and a 118% increase in bytes downloaded for those eleven pages. Sites that reuse even more resources across pages will be more significantly impacted.

Another interesting quirk of Edge’s InPrivate implementation is that the browser will not download FavIcons while InPrivate. Surprisingly (and likely accidentally), the suppression of FavIcon downloads also occurs in any non-InPrivate windows so long as any InPrivate window is open on the system.

Web Platform Storage

Akin to the HTTP caching and cookie behaviors, browsers running in Private mode must restrict access to HTTP storage (e.g. HTML5 localStorage, ServiceWorker/CacheAPI, IndexedDB) to help prevent association/identification of the user and to avoid leaving traces behind locally. In some browsers and scenarios, storage mechanisms are simply set to an “ephemeral partition” while in others the DOM APIs providing access to storage are simply configured to return “Access Denied” errors.

You can explore the behavior of various storage mechanisms by loading this test page in Private mode and comparing to the behavior in non-Private mode.

Network Features

Beyond the typical Web Storage scenarios, browser’s Private Modes should also undertake efforts to prevent association of users’ Private instance traffic with non-Private instance traffic. Impacted features here include anything that has a component that behaves “like a cookie” including TLS Session Tickets, TLS Resumption, HSTS directives, TCP Fast Open, Token Binding, ChannelID, and the like.

Automatic Authentication

In Private mode, a browser’s AutoComplete features should be set to manual-fill mode to prevent a “NameTag” vulnerability, whereby a site can simply read an auto-filled username field to identify a returning user.

On Windows, most browsers support silent and automatic authentication using the current user’s Windows login credentials and either the NTLM and Kerberos schemes. Typically, browsers are only willing to automatically authenticate to sites on “the Intranet“. Some browsers behave differently when in Private mode, preventing silent authentication and forcing the user to manually enter or confirm an authentication request.

In Firefox Private Mode and Edge InPrivate, the browser will not automatically respond to a HTTP/401 challenge for Negotiate/NTLM credentials.

In Chrome Incognito, Brave Incognito, and IE InPrivate, the browser will automatically respond to a HTTP/401 challenge for Negotiate/NTLM credentials even in Private mode.

Notes:

  • In Edge, the security manager returns MustPrompt when queried for URLACTION_CREDENTIALS_USE.
  • Unfortunately Edge’s Kiosk mode runs InPrivate, meaning you cannot easily use Kiosk mode to implement a display that projects a dashboard or other authenticated data on your Intranet.
  • For Firefox to support automatic authentication at all, the
    network.negotiate-auth.allow-non-fqdn and/or network.automatic-ntlm-auth.allow-non-fqdn preferences must be adjusted.

Detection of Privacy Modes

While browsers generally do not try to advertise to websites that they are running inside Private modes, it is relatively easy for a website to feature-detect this mode and behave differently. For instance, some websites like the Boston Globe block visitors in Private Mode (forcing login) because they want to avoid circumvention of their “Non-logged-in users may only view three free articles per month” paywall logic.

Sites can detect privacy modes by looking for the behavioral changes that signal that a given browser is running in Private mode; for instance, indexedDB is disabled in Edge while InPrivate. Detectors have been built for each browser and wrapped in simple JavaScript libraries. Defeating Private mode detectors requires significant investment on the part of browsers (e.g. “implement an ephemeral mode for indexedDB”) and as a consequence most browsers have punted on this problem for the time being.

Advanced Private Modes

Generally, mainstream browsers have taken a middle ground in their privacy features, trading off some performance and some convenience for improved privacy. Users who are very concerned about maintaining privacy from a wider variety of threat actors need to take additional steps, like running their browser in a discardable Virtual Machine behind an anonymizing VPN/Proxy service, disabling JavaScript entirely, etc.

The Brave Browser offers a “Private Window with Tor” feature that routes traffic over the Tor anonymizing network; for many users this might be a more practical choice than the highly privacy-preserving Tor Browser Bundle, which offers additional options like built-in NoScript support to help protect privacy.

 

-Eric

Standard
browsers, security, storytelling

An Update on the Edge XSS Filter

In Windows 10 RS5 (aka the “October 2018 Update”), the venerable XSS Filter first introduced in 2008 with IE8 was removed from Microsoft Edge. The XSS Filter debuted in a time before Content Security Policy as a part of a basket of new mitigations designed to mitigate the growing exploitation of cross-site scripting attacks, joining older features like HTTPOnly cookies and the sandbox attribute for IFRAMEs.

The XSS Filter feature was a difficult one to land– only through the sheer brilliance and dogged persistence of its creator (David Ross) did the IE team accept the proposal that a client-side filtering approach could be effective with a reasonable false positive rate and good-enough performance to ship on-by-default. The filter was carefully tuned, firing only on cross-site navigation, and in need of frequent updates as security researchers inside and outside the company found tricks to bypass it. One of the most significant technical challenges for the filter concerned how it was layered into the page download pipeline, intercepting documents as they were received as raw text from the network. The filter relied evaluating dynamically-generated regular expressions to look for potentially executable markup in the response body that could have been reflected from the request URL or POST body. Evaluating the regular expressions could prove to be extremely expensive in degenerate cases (multiple seconds of CPU time in the worst cases) and required ongoing tweaks to keep the performance costs in check.

In 2010, the Chrome team shipped their similar XSS Auditor feature, which had the luxury of injecting its detection logic after the HTML parser runsdetecting and blocking reflections as they entered the script engine. By throttling closer to the point of vulnerability, its performance and accuracy is significantly improved over the XSS Filter.

Unfortunately, no matter how you implement it, clientside XSS filtration is inherently limited– of the four classes of XSS Attack, only one is potentially mitigated by clientside XSS filtration. Attackers have the luxury of tuning their attacks to bypass filters before they deploy them to the world, and the relatively slow ship cycles of browsers (6 weeks for Chrome, and at least a few months for IE of the era) meant that bypasses remained exploitable for a long time.

False positives are an ever-present concern– this meant that the filters have to be somewhat conservative, leading to false-negative bypasses (e.g. multi-stage exploits that performed a same-site navigation) and pronouncements that certain attack patterns were simply out-of-scope (e.g. attacks encoded in anything but the most popular encoding formats).

Early attempts to mitigate the impact of false positives (by default, neutering exploits rather than blocking navigation entirely) proved bypassable and later were abused to introduce XSS exploits in sites that would otherwise be free of exploit (!!!). As a consequence, browsers were forced to offer options that would allow a site to block navigation upon detection of a reflection, or disable the XSS filter entirely.

Surprisingly, even in the ideal case, best-of-class XSS filters can introduce information disclosure exploits into sites that are free of XSS vulnerabilities. XSS filters work by matching attacker-controlled request data to text in a victim response page, which may be cross-origin. Clientside filters cannot really determine whether a given string from the request was truly reflected into the response, or whether the string is naturally present in the response. This shortcoming creates the possibility that such a filter may be abused by an attacker to determine the content of a cross-origin page, a violation of Same Origin Policy. In a canonical attack, the attacker frames a victim page with a string of interest in it, then attempts to determine that string by making a series of successive guesses until it detects blocking by the XSS filter. For instance, xoSubframe.contentWindow.length exposes the count of subframes of a frame, even cross-origin. If the XSS filter blocks the loading of a frame, its subframe count is zero and the attacker can conclude that their guess was correct.

In Windows 10 RS4 (April 2018 update), Edge shipped its implementation of the Fetch standard, which redefines how the browser downloads content for page loads. As a part of this massive architectural shift, a regression was introduced in Edge’s XSS Filter that caused it to incorrectly determine whether a navigation was cross-origin. As a result, the XSS Filter began running its logic on same-origin navigations and skipping processing of cross-origin navigations, leading to a predictable flood of bug reports.

In the process of triaging these reports and working to address the regression, we concluded that the XSS Filter had long been on the wrong side of the cost/benefit equation and we elected to remove the XSS Filter from Edge entirely, matching Firefox (which never shipped a filter to begin with).

We encourage sites that are concerned about XSS attacks to use the client-side platform features available to them (Content-Security-Policy, HTTPOnly cookies, sandboxing) and the server-side patterns and frameworks that are designed to mitigate script injection attacks.

-Eric Lawrence

Standard
browsers, dev

Streaming Audio in Edge

This issue report complains that Edge doesn’t stream AAC files and instead tries to download them. It notes that, in contrast, URLs that point to MP3s result in a simple audio player loading inside the browser.

Edge has always supported AAC so what’s going on?

The issue here isn’t about AAC, per-se; it’s instead about whether or not the browser, upon direct navigation to an audio stream, will accommodate that by generating a wrapper HTML page with an <audio> element pointed at that audio stream URL.

PlaceholderPage

A site that wants to play streaming AAC in Edge (or, frankly, any media type, for any browser) should consider creating a HTML page with an appropriate Audio or Video element pointed at the stream.

The list of audio types for which Edge will automatically generate a wrapper page does not include AAC:

audio/mp4, audio/x-m4a, audio/mp3, audio/x-mp3, audio/mpeg,
audio/mpeg3, audio/x-mpeg, audio/wav, audio/wave, audio/x-wav,
audio/vnd.wave, audio/3gpp, audio/3gpp2

In contrast, Chrome creates the MediaDocument page for a broader set of known audio types:

static const char* const kStandardAudioTypes[] = {
 "audio/aac",  "audio/aiff", "audio/amr",  "audio/basic",  "audio/flac",
 "audio/midi",  "audio/mp3",  "audio/mp4",  "audio/mpeg",  "audio/mpeg3", 
 "audio/ogg", "audio/vorbis",  "audio/wav",  "audio/webm",  "audio/x-m4a",
 "audio/x-ms-wma",  "audio/vnd.rn-realaudio",  "audio/vnd.wave"};

If the the response sends Content-Type: application/octet-stream, includes a Content-Dispostion: attachment, or puts a download attribute on the anchor <a> element that leads to the media, Edge will download the media file instead of playing it in the browser.

Note: In Windows 10 RS5, the extension model is capable enough that it’s possible to write a browser extension that intercepts navigation directly to audio/video Media types and renavigates to a wrapper page. [Sample code]

-Eric

PS: Edge has similar special handling for video types:

"application/mp4","video/mp4","video/x-m4v","video/3gpp",
"video/3gpp2","video/quicktime"

 

Standard
browsers, privacy, tech, web

Cookie Controls, Revisited

Update: The October 2018 Cumulative Security Update (KB4462919) brings the RS5 Cookie Control changes described below to Windows 10 RS2, RS3, and RS4.

Cookies are one of the most crucial features in the web platform, and large swaths of the web don’t work properly without them. Unfortunately, cookies are also one of the primary mechanisms that trackers and ad networks utilize to follow users around the web, potentially impacting users’ privacy. To that end, browsers have offered cookie controls for over twenty years.

Back in 2010, I wrote a summary of Internet Explorer’s Cookie Controls. IE’s cookie controls were very granular and quite powerful. The basic settings were augmented with P3P, a once-promising feature that allowed sites to advertise their privacy practices and browsers to automatically enforce users’ preferences against cookies. Unfortunately, major sites created fraudulent P3P statements, regulators failed to act, and the entire (complicated) system collapsed. P3P was removed from IE11 on Windows 10 and never implemented in Microsoft Edge.

Instead, Edge offers a very simple cookie control in the Privacy and Security section of the settings. Under the Cookies option, you have three choices: Don’t block cookies (the default), Block all cookies, and Block only third party cookies:

CookieSetting

This simple setting hides a bunch of subtlety that this post will explore.

Cookie => Cookie-Like

For the October 2018 update (aka “Redstone Five” aka “RS5”) we’ve made some important changes to Edge’s Cookie control.

The biggest of the changes is that Edge now matches other browsers, and uses the cookie controls to restrict cookie-like storage mechanisms, including localStoragesessionStorageindexedDB, Cache API, and ServiceWorkers. Each of these features can behave much like a cookie, with a similar potential impact on users’ privacy.

While we didn’t change the UI, it would be accurate to change it to:

CookieLike

This change improves privacy and can even improve site compatibility. During our testing, we were surprised to discover that some website flows fail if the browser blocks only 3rd party cookies without also blocking 3rd-party localStorage. This change brings Edge in line with other browsers with minor exceptions. For example, in Firefox 62, when 3rd-party site data is blocked, sessionStorage is still permitted in a 3rd-party context. In Edge RS5 and Chrome, 3rd party sessionStorage is blocked if the user blocks 3rd-party cookies.

Block Setting and Sending

Another subtlety exists because of the ambiguous terminology “third-party cookie.” A cookie is just a cookie– it belongs to a site (eTLD+1). Where the “party” comes into play is the context where the cookie was set and when it is sent.

In the web platform, unless a browser implements restrictions:

  • A cookie set in a first-party context will be sent to a first-party context
  • A cookie set in a first-party context will be sent to a third-party context
  • A cookie set in a third-party context will be sent to a first party context
  • A cookie set in a third-party context will be sent to a third-party context

For instance, in this sample page, if the IFRAME and IMG both set a cookie, these cookies are set in a third-party context:Contexts

  • If the user subsequently visits domain2.com, the cookie set by that 3rd-Party IFRAME will now be sent to the domain2.com server in a 1st-Party context.
  • If the user subsequently visits domain3.com, the cookie set by that 3rd-Party IMG will now be sent to the domain3.com server in a 1st-Party context.

Historically, Edge and IE’s “Block 3rd party cookies” options controlled only whether a cookie could be set from a 3rd party context, but did not impact whether a cookie initially set in a 1st party context would be sent to a 3rd party context.

As of Edge RS5, setting “Block only 3rd party cookies” will now also block cookies that were set in a 1st party context from being sent in a 3rd-party context. This change is in line with the behavior of other browsers.

Edge Controls Impacted By Zones

With the move from Internet Explorer to Edge, the Windows Security Zones architecture was largely left by the wayside.

Zones

However, cookie controls are one of a small number of exceptions to this; Edge applies the cookie restrictions only in the Internet Zone, the zone almost all sites fall into (outside of users on corporate networks).

Perhaps surprisingly, cookie-like features and the document.cookie getter are restricted, even in the Intranet and Trusted zones.

Chrome and Firefox do not take Windows Security Zones into account when applying cookie policies.

Test Cases

I’ve updated my old “Cookies” test page with new storage test cases. You can set your browser’s privacy controls:

Block3rdPartyChrome

Block3rdPartyFF

…then visit the test page to see how the browser limits features from 3rd-party contexts. You can use the Swap button on the page to swap 1st-party and 3rd-party contexts to see how restrictions have been applied. You should see that the latest versions of Chrome, Firefox, and Edge all behave pretty much the same way.

One interesting exception is that when configured to Block 3rd-party Cookies, Edge still allows 3rd-party contexts to delete their own cookies. (This is used by federated logout pages, for instance). Chrome does not allow deletion in this scenario– the attempt to delete cookies is ignored.

 

-Eric


Appendix: Chromium Audit

In the course of our site-compatibility investigations, I had a look at Chromium’s behavior with regard to their cookie controls. In Chromium, Blink asks the host application for permission to use various storages, and these chokepoints check:

cookie_settings_->IsCookieAccessAllowed(origin_url, top_origin_url);

…which is sensitive to the various “Block Cookies” settings.

Mojo messages come up through renderer_host/chrome_render_message_filter.cc, gating access to

Additionally, ChromeContentBrowserClient gates

Elsewhere, IsCookieAccessAllowed is used to limit:

  • Flash Storage (PP_FLASHLSORESTRICTIONS_BLOCK)
  • Client Hints

Of these, Edge does not support WebSQL, FileSystem, SharedWorker, or Client Hints.

Standard
browsers, dev

ERROR_INSUFFICIENT_BUFFER and Concurrency

Many classic Windows APIs accept a pointer to a byte buffer and a pointer to an integer indicating the size of the buffer. If the buffer is large enough to hold the data returned from the API, the buffer is filled and the API returns S_OK. If the buffer supplied is not large enough to hold all of the data, the API instead returns ERROR_INSUFFICIENT_BUFFER, updating the supplied integer with the length of the buffer required. The client is expected to reallocate a new buffer of the specified size and call the API again with the new buffer and length.

For example, the InternetGetCookieEx function, used to query the WinINET networking stack for cookies for a given URL, is one such API. The GetExtendedTcpTable function, used to map sockets to processes, is another.

The advantage of APIs with this form is that you can call the API with a reasonably-sized stack buffer and avoid the cost of a heap allocation unless the stack buffer happens to be too small.

In the case of Internet Explorer and Edge, the document.cookie DOM API getter’s implementation first calls the InternetGetCookieEx API with a 1024 WCHAR buffer. If the buffer is big enough, the cookie string is then immediately returned to the page.

However, if ERROR_INSUFFICIENT_BUFFER is returned instead (and if the size needed is 10240 characters (MAX_COOKIE_LEN) or fewer), the API will allocate a new buffer on the heap and call the API again. If the API succeeds, the cookie string is returned to the page, otherwise if any error is returned, an empty string is returned to the page.

Wait. Do you see the problem here?

It’s tempting to conclude that the document.cookie API doesn’t need to be thread-safe–JavaScript that touches the DOM runs in one thread, the UI thread. But cookies are a form of data storage that is available across multiple threads and processes. For instance, subdownload network requests for the page’s resources can be manipulating the cookie store in parallel, and if I happen to have multiple tabs or windows open to the same site, they’ll be interacting with the same cookie jar.

So, consider following scenario: The document.cookie implementation calls InternetGetCookieEx but gets back ERROR_INSUFFICIENT_BUFFER with a required size of 1200 bytes. The implementation dutifully allocates a 1200 byte buffer, but before it gets the chance to call InternetGetCookieEx again, an image on the page sets a new 4 byte cookie which WinINET puts in the cookie jar. Now, when InternetGetCookieEx is called again, it again returns ERROR_INSUFFICIENT_BUFFER because the required buffer is now 1204 characters. Because document.cookie isn’t using any sort of loop-until-success, it returns an empty cookie string.

Now, this is all fast native code (C/C++), so surely this sort of thing is just theoretical… it can’t really happen on a fast computer, right?

Around ten years ago, I showed how you can use Meddler to easily generate a lot of web traffic for testing browsers. Meddler is a simple web server that has a simple GUI code editor slapped on the front (most developers would use node.js or Go for such tasks). I quickly threw together a tiny little MeddlerScript which exercises cookies by loading cookie-setting images in a loop and monitoring the document.cookie API to see if it ever returns an empty string.

Boy, does it ever. On my i7 machines, it usually only takes a few seconds to run into the buggy case where document.cookie returns an empty string.

Failure

I haven’t gone back to check the history, but I suspect this IE/Edge bug is at least fifteen years old.

After confirming this bug, it felt strangely familiar, as if I’d hit this landmine before. Then, as I was writing this post, I realized when… Back in 2011, I shared the C# code Fiddler uses for mapping a socket to a process. That code relies on the GetExtendedTcpTable API, which has the same reallocate-then-reinvoke design. Fortunately, I’d fixed the bug a few weeks later in Fiddler, but it looks like I never updated my blog post (sorry about that).

-Eric

PS: Unrelated, but one more pitfall to be aware of: InternetGetCookieExW has a truly bizarre shape, in that the lpdwSize argument is a pointer to a count of wide characters, but if ERROR_INSUFFICIENT_BUFFER is returned, the size argument is set to the count of bytes required.

PPS: As of Windows 10 RS3, Edge (and IE) support 180 cookies per domain to match Chrome, but the network stack will skip setting or sending individual cookies with a value over 5120 bytes.

Standard
browsers

Edge Interop Issues

As we finish up the next release of Windows 10 (Fall 2018), my team is hard at work triaging incoming bugs.

Many such bugs take the form “Edge does the wrong thing for this page. works okay.

This post is designed to be an (ever-growing) index of some of the behavioral deltas that are the root cause of such issues:


Edge doesn’t allow navigation to DATA urls, even when they’d otherwise be converted to file downloads.

Using pushState or replaceState with |undefined| as the URL argument shows “undefined” in the Address box in Edge/IE but not Chrome or Firefox.

IE/Edge strip the Content-Encoding header from a compressed response; Firefox and Chrome leave the header in. For XmlHttpRequest’s getAllResponseHeaders, IE and Firefox maintain the case of HTTP Response header names while Chrome/Edge/Safari do not.

Chrome recognizes that a file with a .JSON extension has the type application/json (and vice versa) while IE/Edge only recognize that when the registry is configured with that mapping.

Chrome includes a hack that works around certificates that do not exactly match the domain on which they are served. Firefox, Edge, and IE do not include this hack, leading to a Certificate Name Mismatch Error when loading:

WWWAddition

Edge does not fully support the URL standard, meaning that URLs of the form http:/example.com (note the missing slash) do not work as expected.

Edge and IE do not allow navigation to HTTP URLs containing a UserInfo component. Other browsers currently (reluctantly) allow this syntax.

Edge RS5 introduces support for Web Authentication specification (in order to support FIDO2 tokens). That specification extends the Credential Management API with new methods, so the navigator.credentials object now exists. However, Edge does not implement the navigator.credentials.preventSilentAccess() method and attempting to call it will cause an exception due to the missing method. (Edge always prevents silent access, so a future implementation of this method will simply fulfill the promise immediately).

When a server returns a HTTP/[301|302|303|307|308] response, Edge/IE are unable to read the response body if the server didn’t include a Content-Length header or Transfer-Encoding: chunked (HTTP/1.1). This turns out to break login to YouTube TV, where Google returns a response body over HTTP/2 (which does not require explicit content lengths thanks to its inherent message framing).

Edge supports most of CSP2 but currently does not support nonces on sourced script elements (only inline script and styles) [Test page]. This limitation significantly complicates deployment of CSP for sites that cannot easily enumerate their source locations in the Content-Security-Policy header (Edge does not support CSP3’s strict-dynamic directive yet either). A broad rule (e.g. script-src https: ) can be used as a workaround but this does increase attack surface.


(…to be continued…)

-Eric

Standard
browsers, security

Stop Spilling the Beans

I’ve written about Same Origin Policy a bunch over the years, with a blog series mapping it to the Read/Write/Execute mental model.

More recently, I wrote about why Content-Type headers matter for same-origin-policy enforcement.

I’ve just read a great paper on cross-origin infoleaks and current/future mitigations. If you’re interested in browser security, it’s definitely worth a read.

Standard