Businesses Urged To Use Unofficial Windows Patch

frankie writes "ZDNet is reporting on the latest dire pronouncements about the WMF vulnerability. The problem is so serious that security experts are urging IT firms to use the unofficial patch. Microsoft's current goal is to release the update on Tuesday." From the ZDNet article: "This is a very unusual situation -- we've never done this before. We trust Ilfak, and we know his patch works. We've confirmed the binary does what the source code said it does. We've installed the patch on 500 F-Secure computers, and have recommended all of our customers do the same. The businesses who have installed the patch have said it's highly successful" It's big enough that even mainstream media is covering the flaw.

Does MS view this as important?

So if this vulnerability is high on the seriousness level, is anyone else wondering the same thing as I am; How and why is it that Microsoft is days behind a third party in releasing a security patch? I mean this is hitting mainstream media, and Microsoft's security patch response team is being bested by some 'guy'?

It brings interesting schemes into my mind. Oh don't mind me, I'm just going to grab my tin foil hat.

Re:Does MS view this as important?

It looks like Microsoft is allowing its user community to patch problems before it can. Oh no!! That sounds a lot like how the Linux community works. Is this going to be a more common occurence as time goes on?

Re:Does MS view this as important?

This'd be a hell of a lot easier if they'd just give over the source code already.

Re:Does MS view this as important?

There is a quid pro quo in the "Linux community". Yes, J. Random Hacker is encouraged (and really expected) to patch Linux flaws. But he recieves a Free system with source code in exchange.

It doesn't sit well with me to see Microsoft eat their cake and have it too.

-Peter

Are you kidding?

This guy (he may be reknowned in the security community, but I've never heard of him) was able to successfully bandage a Windows flaw before Microsoft, without access to the Windows source code or any backing from the writers of the program being patched. I doubt he'll need to look far for work for a long time, and if he does, 'Successfully wrote a patch for a Windows flaw independently' looks damn good on his resume. He still has to pay for Windows, sure, but it's not like he's going to be completely unrewarded for his work.

Re:Does MS view this as important?

This may sound mean-spirited but I think in this case, and any like it, I couldn't blame the security community if it just threw up its hands and said:

"Oh, what a horrible situation -- we could issue our own fix that we've written to help you out, MS -- it's ready to go, we know it works -- but due to the DMCA, Trusted Computing, numerous restrictive MS EULAs and the general legal climate you and other large proprietary software vendors have created, we are genuinely afraid to release our change, as it has required us to disassemble, reverse-engineer and generally do things that you would sue us for. Sorry. Good luck to your *own* patch team."

Why, from a moral standpoint, should anyone help MS do their QA? They certainly have proven themselves willing to sue anyone for any number of reasons relating to reverse-engineering their code -- after all, their philosophy is that no one outside of their teams should know about the OS internals in this way.

They can't have it both ways -- either welcome the users' rights to improve the system they paid for, or don't.

(Yes, I realize that this patch was made to benefit the public in general, and to defend everyone's systems, not directly to benefit MS. But MS does get a free lunch out of this, in some respects.)

MS has to test very extensively

If you're curious as to what all they do, you can take a look here [eweek.com]. A sample quote from the article:

In some cases, particularly when the Internet Explorer browser is involved, the testing process "becomes a significant undertaking," Toulouse said. "It's not easy to test an IE update. There are six or seven supported versions and then we're dealing with all the different languages. Our commitment is to protect all customers in all languages on all supported products at the same time, so it becomes a huge undertaking."

The issue was actually a feature...

According to this F-Secure's Web log [f-secure.com], it tells what is going wrong with the Windows Metafiles (WMF) vulnerability. It turns out this is not really a bug, it's just a bad design from another era. When Windows Metafiles were designed in late 1980s, a feature was included that allowed the image files to contain actual code. This code would be executed via a callback in special situations. This was not a bug; this was something which was needed at the time. The feature now in the limelight is known as the Escape() function and especially the SetAbortProc subfunction, and has been around since Windows 3.0, shipped in 1990...

Seen on Digg [digg.com]. This Broadband Reports' security forum thread [broadbandreports.com] mentioned this as well.

Copied and pasted from my AQFL Web site [aqfl.net].

Re:The issue was actually a feature...

When can I expect a patch for Windows for Workgroups 3.11?

Re:The issue was actually a feature... - WRONG

It's a bug because it doesn't have the .exe extension- if Microsoft tells us "don't download executables from untrustworthy sources" they mean .exe files- they don't mean .jpg files.

Read the Fucking Back Story: This would be almost 0% issue if any of the following were true:

1. MSIE/SHELLDOC used extensions or mime-types (MSIE) in determining what file format something was [[ This flaw is transparent to users: it can be in almost any file extension ]]

2. MSIE/SHELLDOC had a feature like the mailcap file on UNIX which allows us to only list programs that can operate on untrustworthy files(!)

3. The WMF magic was outside of a critical system component (that could simply be unregistered and removed)

As a result, this is a very serious problem, and by playing Microsoft's tune about how "it's not that big of a deal", you're only making the problem worse.

By the way, someone should (quick!) make some WMF files that use the AbortProc routines to disable printscreen and stuff when they're visible so they can sue MS for DCMA (copy protection circumvention) violations...

Re:MS has to test very extensively

Our commitment is to protect all customers in all languages on all supported products at the same time, so it becomes a huge undertaking.

So in other words, we won't release a cure for cancer until we have cures for all other diseases as well.

Re:MS has to test very extensively

No, it wouldn't. That's a bad analogy. Your analogy would more accurately describe a situation where they were sitting on a patch until multiple bug fixes were implemented.

A better analogy would be that Microsoft is withholding the cure for breast cancer until they verify that it doesn't cause patients with other cancers to worsen, that it really does cure breast cancer on more than just one woman, and that it doesn't kill patients outright. with QA, at minimum you've got to verify that a patch can be installed, can be uninstalled if that's an option, fixes the problem, is stable, and passes any baseline usage tests that you have.

The analogy still isn't perfect, but it's far more representative of what a QA process is.

Add the unofficial patch to the test matrix...

This puts MSFT in an interesting position -- their official patch has to be tested on systems with the unofficial patch. Otherwise there's a possibility that the unofficial patch will break something in the official patch (or vice versa.)

With the unofficial patch already deployed on thousands (millions?) of machines, it would be a big deal if something went wrong.

God, I'd hate to be in Redmond right now...

-ch

Bullshit.

Testing?

Even if it means, in contravention of best security practice and all possible "trustworthy computing", knowingly delaying an urgent, critical fix (which would be less troublesome than the first Shatter fix which was pushed out, and only disable a single GDI function that frankly hasn't been used since Windows 3.1 and should never have been used in the first place) for a publically-disclosed, unpatched vulnerability that had been discovered from a 0day exploit, for an indefinite amount of time over a public holiday period while the vulnerability is being "tested"?

When there's realistically no possible way the different L10n's of Windows would affect the GDI32 core because it contains almost no l10n strings anyway, and the vulnerability is in fact a purposely-designed, never-used legacy "feature" that should definitely have been removed in Windows NT or during the Windows 2000 GDI rewrites, or noticed, say, during last months GDI audit?

Despite Microsoft promising that the introduction of the Patch Tuesday would not preclude emergency fixes being issued out-of-cycle and as soon as possible for, ooh, say, critical core Windows vulnerabilities with an enormous number of possible vectors of infection, no effective mitigation and wide, dangerous exploits in the wild with a number of vulnerable machines easily capable of providing an ample breeding ground for supporting wide botnets or enormous worm infections?

Which is exactly what has happened, as Windows has, frankly, just faced the worst single vulnerability in its entire history?*

What the fuck are they doing, deliberately trying to breed another big internet worm?

Sorry, but I'm calling bullshit. I'm a security researcher, and I'm really quite angry at Microsoft's piss-poor handling of this. They couldn't have done much worse if they'd heard about the bug and then have let MSRC take Christmas off anyway.

This was not business as usual. This was an exceptional event (true 0days are actually quite rare to discover in the wild). It could not, and should not, have waited until the next patch cycle. This is exactly the kind of situation upon which a speedy mitigation - hours to days, but definitely not weeks - is absolutely critical, and we should demand that. They should AT LEAST have provided the (untested) hotfix themselves within a day, and pushed it out to Automatic Updates and Windows Update/Microsoft Update within the week after first discovery in the wild - not unrealistic goals for a vendor who wishes to paint themselves as "trustworthy".

They should be brought to task on this one. Behaviour like this is what created the full-disclosure movement in the first place.

* Yes, I'm going to say this one's actually worse than the various active remote vulnerabilities we've had over the years, like the UPnP vuln or the numerous RPC-related vulns. Those, you could at least block with a firewall. This, it's single-payload, multi-vector. It's got plenty of room to drop anything, it's capable of highly metamorphic exploit streams, can be fed online or offline, even spread on media, anything from email to a web page to a simple read-only directory listing or right-click, or uploaded to a site or blog, god help you, rendered inside MSN... the number of potential vectors is so numerous and troublesome it even makes analysis difficult; Windows disregarding filenames and extensions and MIME types and using magic sniffing instead, so you can't even block it effectively using a content-inspecting IDS - that's just the icing on the cake. This is a classic vulnerability, a real ticking Christmas present, a true textbook candidate.

Re:Bullshit.

To me, it's a general problem of redundancy. Filetype should be defined once and only once. If type X has magic prefix Y, then the server should identify the data as of type X using MIME and leave out magic prefix Y. The client can then stuff magic prefix Y onto the file when saving it.

Once-and-only-once is the first and last rule of good programming. The moment any information appears in more than one place, things start to hit the fan.

Re:Does MS view this as important?

Keep in mind that MSfts team must ensure compatibility with hundreds of programs before implementing patches. An independent developer who comes up with a patch doesn't. My 2 cents.

Re:Does MS view this as important?

I wouldn't call it hundreds.
Even so, it probably just a few code libraries to check against as I doubt they check against each and every title listed here:
http://support.microsoft.com/gp/lifeselect [microsoft.com]

Probably their main concern is the Enterprise level support they have to comply with and NOT rush a patch out.

Re:Does MS view this as important?

Presumably they do some sort of testing with their patches before they release...

Re:Does MS view this as important?

"What's the liability for the 3rd party if their patch screws something up in a bad way? Zippo. That's (part of) the reason why it takes longer to put out an "official" patch."

What's the liability if MS screws up a patch? They do it all the time, but I don't hear anything about them being sued or compensating businesses they've hurt.

Re:Does MS view this as important?

That would be the same as the liability that Microsoft would have if its patch screwed something up, right? Zippo in either case. RTFEULA.

The problem is it's a GDI exploit

The actual root of the problem is in the GDI, which is what handles all basic interface display for Windows. The unofficial patch just disables the call that the exploit uses. Ok, fair enough, but that's a hack, not a fix. That means that anything that legitmately uses that call won't work, and the underlying problem is still there.

Well, testing a fix for a system component like that takes time, espically since it affects a ton of versions.

Now you might ask, why not release a hack fix, and then do a proper patch later? Well as it stands, it's hard enough to get people to update their systems. We fight with it all the time with people here at work. They turn auto updates off since they run simulations at night and don't want it rebooting (even though patch day is known ahead of time) and then never manually patch since they "can't be bothered".

Well, if MS released a patch that broke things, that just makes that many more people stop patching. Remember all the whining and bitching about SP2. There were very few systems that had problems with it, and most that did were spywared to hell, but still there are tons of people that refuse to install it for fear that "it'll break my computer".

Thus the offical patch takes time, as they have to test and make sure that the problem really is fixed, and no new problems were created with the fix. REgression testing isn't quick.

Re:block wmf

Its not just the extension that dictates that it's a WMF... Windows in its infinate wisdom also looks at the header bytes of the file and says "ohh! thats a WMF!" Execute! im in a damned hurry, hopfully I stated that correctly...ymmv

Re:block wmf

Its not just the extension that dictates that it's a WMF... Windows in its infinate wisdom also looks at the header bytes of the file and says "ohh! thats a WMF!"

So, in other words, it does exactly the same thing Unix does for every single executable file.

Do a man magic if you don't know what I'm talking about, and/or look into why scripts have that #! as the very first two bytes in order to work automatically.

Windows has gotten bashed for years for relying on file extensions. Here they don't and they get bashed more! Ok, yeah, it's yet another example of deviation from expected behavior, but complain about that, not that they're finally trying to be smarter about files. Hell, most programs will now ignore file extensions and look at the file header -- it's hardly a MS only behavior.

That said, MS's slackness on this issue is ridiculous. Yes, I know that they have to test a patch in a very large test environment to make sure nothing goes "boom", but in this case they would better serve their customers by simply disabling WMF support entirely until they can properly patch things. WMF is not a widely used format -- in the very few cases where it's actually being used you could simply not patch the computer and take appropriate actions to isolate that system. It would be a hell of a lot better than the current situation, especially given how nasty and widespread this exploit is.

Re:block wmf

That's great, but it's all irrelevant. The HTTP 1.1 protocol says that a browser shouldn't try to guess the MIME type of a document if it's specified by the server. IE ignores this and tries to guess the MIME type anyway.

Note the key difference between an OS (your example) and a browser (reality).

More details

Vulnerability in Graphics Rendering Engine Could Allow Remote Code Execution

http://www.securityfocus.com/bid/16074 [securityfocus.com]
http://www.microsoft.com/technet/security/advisory /912840.mspx [microsoft.com]
http://www.symantec.com/avcenter/venc/data/pf/pwst eal.bankash.g.html [symantec.com]

WooHoo 3rd parties!

We don't see 3rd parties doing patches for MS problems much :-) They joining the Open Source bandwagon yet?

Ha, so much for such "features" - times have changed...

--LWM

Re:F-Secure are publicity sluts

Huh? How'd get this modded Insightful? It's pretty much the opposite, actually - considering that F-Secure is in the business of security solutions, it's *expected* of them to uncover new problems, and I at least think it's *GREAT* that they decide to make the information available to everyone instead of just rolling it into the next update for their enterprise products.

Think about it - they're doing good research, AND they're making it available for free, and you still criticise them for exactly that? You're not just looking the gift horse into the mouth, buddy, you're trying to paint the giver in a bad light for attempting to give it to you for free.

Seriously, get a grip.

One Gets the Feeling...

One gets the feeling that the MS programmer didn't want to come in over the New Year's holiday to work on some piece of legacy code from 1990 that he was handed several years ago when the last programmer whose responsibility it was, was promoted/left for Google. This latest programmer has never looked into this code before this last weekend.

It may not have been anything like this at all, but this is the feeling one gets.

One also wonders about the job security of the MS programmer who didn't get this fix out in a timely manner.

This is slashdot, wheres the pictures?

Its ok, I found th...!&^!")NO CARRIER

Re:This is slashdot, wheres the pictures?

No Spot! Don't Chew on the power*ZAP!* %^@!NO TERRIER.

Sorry, had to do that. ^.^

The Business Mindset

Businesses are only going to respond to a problem by calling on the person/entity that is supposed to cover it, i.e. the one they're paying, Microsoft, in this case. They're not going to go around installing an independent patch willy-nilly on dozens of computers if it takes another day to get it from Microsoft. Many of these are small businesses without IT departments to advise them one way or the other. The important point here is that by waiting the extra day, a few of them are going to get burned badly and Microsoft will lose much of their trust.

MS workaround

The current official suggestion from MS is to limit problems is of course to unregister the related driver, shimgvw.dll.

Whoa, that's really bizarre

This article isn't anything like the one that I submitted.

• 2006-01-03 17:15:05 No Microsoft WMF update until next week (Index,Windows) (accepted)

Mine looked more like this (body content from memory):

" The usual suspects [google.com] are reporting Microsoft's latest announcement about the WMF vulnerability (link to previous /. article). To quote (link to MS technet article): "Microsoft's goal is to release the update on Tuesday, January 10, 2006, as part of its monthly release of security bulletins." So do you install the unofficial patch (link to previous /. article), or cross your fingers for a week?"

Re:Whoa, that's really bizarre

Just further shows that the "editors" don't even "get" their roles as editors. Attributing words that weren't written to the submitter is not something they should be doing. Or if they do, they should use the standard square brackets to indicate that those words weren't said, but were what was implied. Changing the title is fine. Adding additional commentary or extra sources (as Zonk did with the 'From the ZDNet article' bit) is fine. Putting words in people's mouths is a HUGE editorial no-no.

What will be especially interesting...

will be to compare the Microsoft released patch to the unofficial one.

It would be deliciously muddying for Microsoft if someone discovered significant parts of the unofficial patch in the official one.

Not good enough...

Not all WMF files have the .wmf extension. Some may have .bmp, .gif, .jpeg, or about a dozen others.

I saw a list a few minutes ago, but I don't remember where...

Oblig. Star Trek

Kirk: Fix the WMF hole!

...

Let me guess: Tuesday?

investigation?

Microsoft (Research) said in a security bulletin on its Web site, "we are working closely with our antivirus partners and aiding law enforcement in its investigation."

Cool - law enforcement is investigating Microsoft? About time!

get a rope!

This really IS as bad as SANS says...

Worse, in fact. There are SEVERAL ways, all well known, which could leverage this exploit to compromise millions of hosts in a matter of hours.

The unofficial patch is 100% necessary. This is BAD folks.

And if the evil people are smart, they'd have a very VERY nasty suprise come monday, when most people are still not patched and M$ hasn't released the official patch yet.

Download

If you want the patch itself, try here:
http://isc.sans.org/diary.php?storyid=1010 [sans.org]

Second time this story came up with no links to the patch.