Apple Silicon Hardware: SPRR and Guarded Exception Levels (GXF)
blog.svenpeter.dev
blog.svenpeter.dev
If I understand correctly, one of the big unsaid implications of this article seems to be that it is now impossible to have RWX, because the hardware completely prevents it. APRR looks more like a general permissions-remapping mechanism, while SPRR appears to go beyond that.
While the technical details are certainly interesting, IMHO it's also disturbing in the same way as weapons of mass destruction or DRM. This technology is used by one company to essentially maintain control over its users, under the convincing guise of security.
What do SPRR and GXF stand for?
Given the functionality, my guess is APRR = Access Permission Remapping Registers, SPRR = Secure Permission Remapping Registers, and GXF = Guarded Execution Feature.
Its not a guise of security, it actually accomplishes it. Mac malware now has to operate within the confines of these permissions and nowadays the author has to submit a sample of the malware to Apple before it can be opened without insane prompts (this also means losing $99 and attaching their legal identity to the malware/committing identity fraud). If Apple went full-iOS on MacOS and made it so "everything must go through the App Store" then the possibility of mac malware would effectively disappear overnight into the same realm that iOS malware currently resides: jailbroken iOS devices with shady (mostly piracy-focused) apt repos.
Realistically, malware only affects a tiny fraction of users, and only a further tiny fraction suffer demonstrable loss from it.
Their whole design is incredibly neat and well done! If you like these features (or just don't care) the default install does make attacker's lives harder.
But if you disagree with these features or just don't like them you can just boot into recovery mode, authenticate with your password and disable almost everything for macOS.
And if you just like the hardware you can do the same and install a custom kernel like Linux or *BSD and do whatever you want.
You can even have triple boot into one macOS with full security enabled, another macOS install with everything disable and a third "macOS" which actually is Linux.
They spend a lot of effort and engineering time to make all this possible.
For a certain generation of adults, going to visit your non-technical parents socially meant budgeting extra time to eradicate the adware of the month. It was a constant problem affecting large swaths of users, and fixing it was beyond the technical ability of most people.
Edited: fix typo.
you can also just map the same physical memory once as rw- and once as r-x even with SPRR enabled there.
We just let it be locked in the secret document vaults at Apple, and then had to waste effort rediscovering it.
That doesn't seem smart.
0: https://www.worldofcoca-cola.com/media-alert/coca-cola-moves...
Yet we see the equivalent with humans all the time. Globally less efficient use of time/money/resources for 'fairness'.
No right answer, just curious
For me the basic value is just having more knowledge published in the open for anyone to read. I might not have any practical value of it, either ever or today, but I did learn a bunch reading it and I'm glad I did.
To me, it's a combination of various things:
I've tried to outline the process I used so that other people can learn how to approach challenges like this. I learned how to develop software and how to reverse engineer by reading posts like this and watching talks from e.g. CCC or DEFCON. Hopefully I can give something back this way.
This might also be valuable to other people who want to virtualize macOS or iOS or those who do security research and run into these features while reading XNU code.
And then ofc I can finally flex in front of my friends with this because I was the first to pull of a new skate trick ;)
Also, rad trick ;)
The result would be similar to the protectionist trade policies that hamper market entry for foreign produced goods in India and some Latin American countries like Argentina. There are enough customers elsewhere whose governments wouldn't give a thought to this that non-CA/US companies would safely ignore CA/US in this case. It would only work with coordinated effort.
There are also other second order effects not considered by a policy like this.
What's the timeline and what's the company scale that this levies?
Is the documentation expected on day 1? Great, Apple and Google and Sony and others will squash all small competitors. You will never see another small (< 20) person hardware startup (not that we see them often anyways) because the cost to produce this kind of documentation is non-zero, and the big players can easily absorb it into their processes. Additionally, small players would literally be handing the design to competitors. OK, maybe patents and things like that stop some companies from being bad actors and reproducing the product in short order, but it won't stop everyone especially in countries with laxer or no enforcement of patents.
Additionally, problems would be created if designs were produced in non-compliant countries. If I start up a hardware company and make a new device, but half my components are designed in Taiwan and they've decided to explicitly reject this, am I non-compliant for not releasing documentation I don't have the authority to release? And if I'm able to become exempted because of a 3rd party component, everyone else would start shifting their design departments to contractors (if necessary) in those countries or shift a portion of their company there.
And if you create exemptions for small businesses, the big businesses would do exactly what they always do, find loopholes to get subsidiaries/contractors to do the work and exempted and nullify the entire thing.
As nice as it would be to bring back (easier) access to system specs, it's impractical without fundamental culture changes that laws themselves don't usually produce.
The U.S. regularly imposes plenty of requirements that are far more difficult to meet, in many industries. The U.S. is the largest market in the world for many goods; I think the parent greatly underestimates the power and practices of national governments. Smaller countries also impose requirements.
> You will never see another small (< 20) person hardware startup (not that we see them often anyways) because the cost to produce this kind of documentation is non-zero, and the big players can easily absorb it into their processes.
I highly doubt that small players would be affected at all, because a small player wouldn't produce a custom chip. The level of documentation required is akin to API documentation that many services produce and this is not revealing any trade secrets, but enables customer to make full use of the device. If you create a product without any documentation, then you have bigger problems.
> If I start up a hardware company and make a new device, but half my components are designed in Taiwan and they've decided to explicitly reject this, am I non-compliant for not releasing documentation I don't have the authority to release?
This is false, because it is companies like Apple that tell component manufacturers to not sell or to not disclose documentation. Virtually any chip that you can get on an open market has complete documentation available. If you are ordering a custom chip, you shouldn't be allowed to tell manufacturer to not sell it to anyone else nor hide its documentation if it goes into your product.
If you were actually working in hardware design space you would know that what is being asked does not compromise trading secrets, it is done solely out of greed so that people are nudged towards subscription models and becoming dependent on a particular corporation.
This is just a case of Apple being careless. We have a right to criticize the largest company in the world, if not hold them to the highest standards possible. Instead, we've let Apple off the hook because nobody wants to stand up to them and call them out on it. We're all willing to bury those experiences, as long as our next laptop can export our Photoshop projects 500ms faster.
It is a practical requirement in the sense that it improves (potentially) the practicality of hardware systems.
Those are two distinct senses, I was writing about the former.
Isn't this statement a bit of a blanket? Not too long ago I purchased a laptop specifically for being tiny and lightweight (X1 Nano). Ports and expandability were an afterthought. Certainly many wouldn't mind increased thickness and bulk but there is absolutely a market for portability, at least in laptops and other mobile devices.
That said, the X1 Nano at least has removable/upgradable storage, which can't be found on heavier MacBooks, so there's absolutely an argument for adding expandability where it doesn't negatively impact portability.
I think Apple have a pretty good idea about marketing, and what sells a product and what most customers find attractive about it. The vast majority of computers sold are never upgraded, whether they can be or not and most people would never even think about it.
That's an unfortunate choice of words. I'm definitely looking for a product, not a service (which is why I bought my MacBook instead of leasing it), but everything I need the machine for overlaps with what Apple envisions me doing with the machine.
I don't want my machine to be amateur-repairman friendly and my OS to be infinitely customizable and tinkerable, because we already have that, it's called Linux on a ThinkPad and to me that's the antithesis of a Mac and I consciously decided not to use that kind of setup.
(I'm not strongly for or anti right-to-repair, but I tend to be against tight regulation when avoidable.)
Edited to add: the author doesn't seem to themselves be heartbroken. This level of detail suggests enthusiasm.
Absolutely not! Reverse engineering is a lot of fun to me! Especially the moment of clarity when all the unknown pieces eventually fall together and I finally get the whole picture.
Understanding IC's to repair them is a failing prospect. They are black boxes by nature. If you can't actually repair it without the factory, why should they write you a manual to help you repair it?
Beyond that though, there is a degree of Power lost through dispelling the blackbox (information asymmetry is Power, never forget). Explaining how it works opens up PR doors places would rather not have to deal with, and cuts off avenues of development that companies in the space thrive on. See Nvidia graphics cards for an example. Nvidia doesn't "sell" you the card, in the sense of allowing you to do whatever you want with it because they can't. They are contractually required to hide implementation details around things like HDCP and the in hardware foundations of many streaming related DRM technologies.
Then you have other aspects, like their move to requiring Nvidia signed firmware to access the power control of modern cards. This ensures only their driver can feasibly compete in the market and locks out any potential competition that may be able to pull off a true innovation using their own hardware platform (see nouveau and the signed firmware issue).
All of this would be undone by actually being required to support user firmware implementation, and disclosing how everything works.
https://docs.nvidia.com/drive/drive_os_5.1.6.1L/nvvib_docs/i...
The above is a link to an offering for a self-driving processing offering (therefore safety-criticality somewhat justifies the absurd guarantees of firmware lockdown), but my experience tells me that once ypu nail down a tricky process like that, you worm it into as many parallel business vehicles as you can.
Just knowing another company is even capable of doing something like that is a signal of demand to competitors. Hence the desire to hide everything possible as no one else's business.
All they do is lock out customers from owning their devices. It's about control- plain and simple.
You have two mechanisms to protect from copying: trade secrets and patents.
If you document it, it’s not a trade secret anymore. If another person figures it out, it’s also not a trade secret anymore.
And Apple (and others) document plenty of stuff in patents. However, those aren’t loved by this community either.
Why do you (and several others on HN) keepi making this claim with so much confidence? Do you work in the electrophoretic display industry? Do you know what the costs are to make electrophoretic displays? Is there any evidence that patents are what dominate the costs of producing electrophoretic displays? Or do you think there is even a slight possibility that the dominant factor of price is actually volume and scale of production?
Now finally a few years ago their patents finally lapsed, forcing them to finally compete with other companies on price and quality.
End result - affordable 13 inch pen displays available at about third to quarter of the old Wacom price from multiple vendors.
I can only imagine the situation is similar if the original E-ink developers managed to patent all the necessary technologies to bloc everybody else.
Same thing for 3D printing - it only really took off once Stratasis could no longer milk their FDM patents to block all progress in affordable 3D printing area.
All this imagination is amazing to me as a person who actually works in the display industry. It is like someone walked into a sports car showroom and announced that these cars are too expensive because of patents and it has nothing to do with volume and actual cost of iron and rubber. And then a bunch of other people all agreed with them and started blogging about it. Because that's what I see happening here. I wonder when someone like this will walk into an Apple showroom and start exclaiming that these apple products are all too expensive because of patents and not because of other factors.
As for Wacom and Stratasis, I have no idea since I don't work in those industries. But I've never seen convincing proof put forth. Circumstantial sure, but clear data that would convince me that "Stratasis used FDM patents to block all progress in 3d printing" doesn't seem to be available.
People get irately angry when a warranty replacement ends up with a slightly different system (CPU stepping, firmware revision, or whatever) that breaks something. They get only slightly less angry when buying a newer system does the same: "XYZ worked on my previous system!!!!1 this is broken garbage!!1111 Widget Inc is deliberately screwing over loyal paying customers to force us to buy more widgets, it makes me sick!!!!!1111" - this is not too far off from comments made here on HN at times.
Making seemingly trivial changes becomes an exercise in walking a minefield of unknown compatibility constraints.
That's besides the extra work of hiring technical writers to spend time converting a jumble of engineering notes, comments on bug tickets, and code comments into publicly presentable documentation - and making sure to keep all of that up to date when anything changes. For something almost no one knows or cares about that you may completely change in the next version.
Well, but at least some people at Apple do. Kind of. They built a mechanism for loading alternate kernels while knowing full well that it would require reverse engineering in order to use. A former engineer said as much. https://twitter.com/XenoKovah/status/1339914716454526979
In a way, Apple has done this for years. Their Mail app contains a third-party plugin API (which dates back to at least Tiger, so it's not some unfinished feature they forgot to remove), but it's completely undocumented. You can't use it unless you reverse engineer it.
I don't quite get it, really.
And if you buy anything nowadays you will get undocumented locked-down hardware. It doesn't happen if it's recently invented or existed for centuries.
Compare 1980's sound systems to today's sound systems. Or 1980's fridges to today's fridges. Or washing machines, circuit breakers, whatever.
If you bought a $30000 server, you'll get schematics and a lot of information of the product. Including motherboard traces and gerber files:
Intel have done some really shady things, and yet they still publish extensive documentation on how their processors work and how to get the best from them - Apple? Nothing.
One way to look at it is as ('at it is as'? English is weird) general purpose computer vs. appliance. Apple probably feels no more obligated to give you schematics, chip internals, etc. than Garmin does for a GPS.
Another angle is that hardware companies are becoming more secretive over time. There weren't many secrets in a Marantz receiver or a shortwave radio. Perhaps there's been a change in approach due to low-cost Asian manufacture and the tendency to ignore IP.
There are only so many way ways to skin a cat (in nuts and bolts engineering). When constrained by physics, everyone is going to converge on the same fundamental designs. You therefore have two or three options. Take or leave the licensing agreement of the IP holder, which need not be reasonable or effective; say screw it, use the IP and hide the schemata behind enough trade secret bs red tape that no one is comfortable leaking it, or give up.
I wager the second option contributes to things when number 1 fails. At least, it's the only reason that holds water for me, then again, I haven't gotten into that industry enough to vouch for the presence of skeletal remains in the closet.
I also think that, historically, electronics were documented because they broke down so often.
You just had to be able to spot a broken tube in your radio, clean a potentiometer, etc.
In general, I agree that the world would be a better place if corporations would release documentation.
But my life would be much less fun then. Reverse engineering scratches a certain itch like nothing else does! So I'm kinda happy that mysteries like this exist and gladly spend my time on something like this :-)
Totally, yet at the same time I'm worried that knowledge workers aren't aware enough of just how privileged they are to have been able to 'climb the ladder' to be able to do work like that, and how, unless you're 1) well off and 2) living in the global north, gaining such skills has been made nearly impossible because the ladders needed to climb up are often not available, despite humans having been gifted digital technology (and therefore a zero-marginal cost of information reproduction).
As an example: I'm not technically skilled enough to reverse engineer the things I'd like to, and because of the way knowledge is controlled by big companies as trade secrets and patent claims [1], it's near impossible to gain these skills in an affordable way. I also cannot take apart the technology I already own since it is not modular, it voids the warranty or is it is damaged (since I'm not skilled because I didn't realize early enough how important technology is).
So yes, it sounds super fun for you to reverse engineer this stuff, yet when are we going to seriously admit that this IP system has become a massive problem, and that the people who came before are literally 'kicking away the ladder' they used to climb up themselves? That when people say that the only thing people need to do to succeed is to 'work hard', that that is a lie, since information is made artificially scarce by human systems/institutions, and thus only available to a select few [2].
Edit: I just saw on your twitter that you are a console hacker, thanks for the awesome work!
[1] https://www.ipwatchdog.com/2019/02/19/dont-fooled-patent-pur...
And, to be clear, I do think the current situation of closed systems is not great, and I do think that we do have a lot of engineers who grew up on open systems that they could tinker on going on to design things like iPhone for their children to use. These are real problems, but I just wanted to say that blog posts like these are not the problem; in fact I think they are beneficial as they allow more people to have access to this kind of information.
yup, that's how I learned all of this!
> And, to be clear, I do think the current situation of closed systems is not great,
Agreed!
> and I do think that we do have a lot of engineers who grew up on open systems that they could tinker on going on to design things like iPhone for their children to use.
These days there are also more open system like, say, the Raspberry Pi. Back in my day (oh god, I'm growing old!) we had to first exploit video game consoles to get something comparable :-)
We made a slightly related argument when we didn't really feel like driving the WiiU homebrew scene almost 8 years ago (stop making me feel old!) [1]
* for the privileged! The digital divide is very real.
But I'm not sure realize where the privilege actually is. People aren't "gifted" with technology. Oftentimes it's because they were exposed to it and didn't have other obligations combined with a mixture of curiosity that was triggered by something. I knew people that had to think long and hard if it's worth saving up 150 dollars to buy a laptop or PC to MAYBE learn skills that MAY help them or if they should just continue working as a hostess. If you have to spend 12 hours a day to grind through school and after-school activity and then homework when society has decided that you can't be a productive member of society later AND then have to help you parents on the farm before you collapse of exhaustion then yes, you won't be able to acquire these skills.
I was privileged enough to have my mother sacrifice everything(her property, her life, her family and her homeland) to pull me out of a warzone to provide for me so I would be in a position to be able to build these skills. Just calling the whole thing privilege is such a disservice to all she had to sacrifice for me to be able to be in this position.
If we want to talk about privilege how about the fact that if the garbage worker wouldn't collect our garbage and the farmers wouldn't provide us with food then we wouldn't even be able to do the things we are able to do and while they feed us so we can build the things that replace them, we tell them how they should have just got to school to fill out excel charts, powerpoints and the next world change ai pipeline, or photo sharing app that nobody needs.
Pretty much. I learned how to reverse engineer back when I was still in school and was bored with my Wii. I wanted to run my own code on it to e.g. watch movies on my TV. But I couldn't afford to buy another computer connected to my TV. So I just had to figure out how the Wii works and how I could run my own code on it.
The privilege I had was that my parents were well-off enough to allow me to have >8 hours of free time after school and that they bought me a computer and gave me pocket money for my Wii.
I used the words "humans having been gifted digital technology" as I am a marxist and I believe that all the technologies and scientific knowledge that we reverse engineer from nature is gifted to us by a higher power. Others talk about the brilliance of human minds, yet since I see us as a part of nature, there's no distinction in that for me. Another reason I say it that way is because unfortunately the 'great man theory/myth' still reigns supreme in the tech industry [1]. I think though that this 'brilliant mind' story that is commonly told in the west seems to have often justified detrimental and destructive human supremacy/domination over the rest of nature (also leading us to the edge of the cliff now with global warming).
Anyways, I think that we all tell ourselves different stories and this is one of mine. This is another story I tell myself because it is ridiculously beautiful to me: http://www.galactanet.com/oneoff/theegg_mod.html or Kurzgesagt version: https://www.youtube.com/watch?v=h6fcK_fRYaI
> I was privileged enough to have my mother sacrifice everything(her property, her life, her family and her homeland) to pull me out of a warzone to provide for me so I would be in a position to be able to build these skills.
That sounds incredibly difficult and distressing for her (and possibly for you too). Thank you for sharing this personal story here.
> Just calling the whole thing privilege is such a disservice to all she had to sacrifice for me to be able to be in this position.
That isn't what I am referring to as the 'privilege' here though.
I am saying that the property system is the problem. The system makes it so that there are very few knowledge workers. Big capitalist firms, together with the capitalist nation state, control where and when research and development is to be done, and by whom (by enclosing/owning the systems that control the flow of information).
So I have no doubt that your mother/caretaker(s) sacrificed a lot. No doubt about that at all.
Professor Jakob Rigi does a good job of describing what I am trying to explain, maybe I will let him do the talking. Excuse this long quote by him yet I hope the clarity of his message makes it worth it.
“Digital piracy and the digital copying of cultural products for private use is a refusal to pay rent-tribute to knowledge capitalists. Therefore, piracy is miss-naming of the phenomenon. The sea pirates take away by force others' properties. The digital “pirates” only use universal commons which have been artificially fenced off. They just remove fences, and by doing so they do not take away knowledge, because, knowledge cannot be taken away. They use something which by its nature belongs to the whole of humanity. The producer of knowledge uses knowledge, as “raw” material, which is part of the general intellect of humanity as a whole and the produced knowledge itself becomes immediately part of this general intellect. Therefore, the fencing of knowledge is, essentially, more similar to the traditional piracy. The knowledge capitalist fences off, with help of the force of law, universal commons that does not exclusively belong to her/him. Therefore, s/he robs commons.
To put it bluntly, digital piracy takes back that which has been stolen from the public. Therefore, although illegal, it is morally and ethically justified. The very fact that public ethics and the bourgeois property rights contradict each other on this matter evidences that such rights are superfluous in our era of digital technology. In this way, the digital piracy and digital counterfeiting is an important economic-social movement of our time.
This movement is expressed in various ways including the following. First millions of individuals around the world, understanding and believing that they are not involved in theft, copy things for individual uses. The historical, cultural and political significance of this practice can hardly be exaggerated. It undermines the moral and ethical legitimacy of the bourgeois intellectual property in the very pours and veins of everyday life. Digital piracy is a major force of the growth of knowledge and culture, on the one hand, and the self-improvement of the individual on the other. Second, “pirate” activists, so-called crackers, illegally copy fenced off knowledge and make it available for a global public on the net. A good example was Gigapedia digital library on the net, which was created by activists who scanned books. These activists are either from poorer countries or classes or our era’s Robin Hoods from privileged countries and classes. Aaron Swartz was one such Robin Hood. The very massive and online and off line protests against SOPA in the USA and ACTA (Anti-Counterfeiting Trade Agreement ) in the European Union, and their temporary success, are evidence of the moral legitimacy of digital piracy and digital counterfeiting." [2]
Or as another author puts it:
"The current political economy is based on a false idea of “immaterial scarcity.” It believes that an exaggerated set of intellectual property monopolies – for copyrights, trademarks and patents – should restrain the sharing of scientific, social and economic innovations.3 Hence the system discourages human cooperation, excludes many people from benefiting from innovation and slows the collective learning of humanity. In an age of grave global challenges, the political economy keeps many practical alternatives sequestered behind private firewalls or unfunded if they cannot generate adequate profits." [3]
So where does that leave us? I believe that the most powerful leverage point we have available to us is being explored by the http://valueflo.ws project, in tandem with the http://metacurrency.org project. Those projects meet in the middle in a third project called hREA or holoREA [4]. I've written about them in previous comments:
https://news.ycombinator.com/item?id=25845914
https://news.ycombinator.com/item?id=26587791
[1] https://www.technologyreview.com/2015/08/04/166593/techs-end...
[2] https://www.triple-c.at/index.php/tripleC/article/download/4...
[3] http://wealthofthecommons.org/essay/peer-peer-economy-and-ne...
Some obscure HPC stuff, vintage mainframes, old games with long lost source code, historic military computing harware, CNC controllers & industrial SCADA devices, consumer device MCUs, advanced radio hardware - lot of stuff to dive into without really knowing what to expect. :)
Not at all the case for other kinds of hardware, or any software.
Companies being required to release documentation would increase the cost of development (passed on to the consumer) by a non-trivial amount.
As it is, people inside companies usually cannot get good documentation for the things the company develops that they need to interface with. Been there, seen that. It’s unfortunate, because that can also increase costs, but usually people just muddle through with poor documentation and many email chains to ask questions about the things that are poorly documented.
It’s the sort of job you’d give an intern if you couldn’t find something else for them to do.
In fact, I’d hazard a guess that the large majority of compute capable devices you own, you do not “own” according to your definition of ownership.