Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
Bug Security Intel Operating Systems Software Windows Linux Technology

'Kernel Memory Leaking' Intel Processor Design Flaw Forces Linux, Windows Redesign (theregister.co.uk) 416

According to The Register, "A fundamental design flaw in Intel's processor chips has forced a significant redesign of the Linux and Windows kernels to defang the chip-level security bug." From the report: Programmers are scrambling to overhaul the open-source Linux kernel's virtual memory system. Meanwhile, Microsoft is expected to publicly introduce the necessary changes to its Windows operating system in this month's Patch Tuesday: these changes were seeded to beta testers running fast-ring Windows Insider builds in November and December. Crucially, these updates to both Linux and Windows will incur a performance hit on Intel products. The effects are still being benchmarked, however we're looking at a ballpark figure of five to 30 per cent slow down, depending on the task and the processor model. More recent Intel chips have features -- specifically, PCID -- to reduce the performance hit. Similar operating systems, such as Apple's 64-bit macOS, will also need to be updated -- the flaw is in the Intel x86 hardware, and it appears a microcode update can't address it. It has to be fixed in software at the OS level, or buy a new processor without the design blunder. Details of the vulnerability within Intel's silicon are under wraps: an embargo on the specifics is due to lift early this month, perhaps in time for Microsoft's Patch Tuesday next week. Indeed, patches for the Linux kernel are available for all to see but comments in the source code have been redacted to obfuscate the issue. The report goes on to share some details of the flaw that have surfaced. "It is understood the bug is present in modern Intel processors produced in the past decade," reports The Register. "It allows normal user programs -- from database applications to JavaScript in web browsers -- to discern to some extent the contents of protected kernel memory. The fix is to separate the kernel's memory completely from user processes using what's called Kernel Page Table Isolation, or KPTI."
This discussion has been archived. No new comments can be posted.

'Kernel Memory Leaking' Intel Processor Design Flaw Forces Linux, Windows Redesign

Comments Filter:
  • by Anonymous Coward on Tuesday January 02, 2018 @05:44PM (#55851737)

    At one point, Forcefully Unmap Complete Kernel With Interrupt Trampolines, aka FUCKWIT, was mulled by the Linux kernel team, giving you an idea of how annoying this has been for the developers.

    • by Hal_Porter ( 817932 ) on Tuesday January 02, 2018 @07:51PM (#55852477)

      https://www.mail-archive.com/l... [mail-archive.com]

      2) Namespace

            Several people including Linus requested to change the KAISER name.

            We came up with a list of technically correct acronyms:

                User Address Space Separation, prefix uass_

                Forcefully Unmap Complete Kernel With Interrupt Trampolines, prefix fuckwit_

            but we are politically correct people so we settled for

              Kernel Page Table Isolation, prefix kpti_

            Linus, your call :)

      LOL!

    • At one point, Forcefully Unmap Complete Kernel With Interrupt Trampolines, aka FUCKWIT, was mulled by the Linux kernel team, giving you an idea of how annoying this has been for the developers.

      Someone on LKML using the word "fuckwit" is what Linux developers call "a normal Wednesday".

  • FOOF (Score:5, Insightful)

    by OverlordQ ( 264228 ) on Tuesday January 02, 2018 @05:44PM (#55851739) Journal

    About par for Intel's course. Make it fast at the expense of horrible bugs.

    • not you. thats's what the Linux team wanted to call this bug.

      I read the El Reg article but I still don't understand what it is saying. At all levels. I don't understand if this means all intel processors or just the new ones. I don't understand if the 20% slowdown is for a tiny fraction of operations in the OS or if it means that things like e-mail, firefox or general python programming will be slowed down 20% overall. The latter would be a disaster. (could I ask intel to refund 20% of my computer cos

  • In all fairness... (Score:3, Insightful)

    by Anonymous Coward on Tuesday January 02, 2018 @05:45PM (#55851749)

    Intel guys are doing the bulk of the work for the linux kernel changes, and I'm sure to be fair they'll equally cripple all processors with the changes not just their own.

    • by scumdamn ( 82357 ) on Tuesday January 02, 2018 @05:56PM (#55851847)
      Looks like you missed this commit [lkml.org] from Tom Lendacky at AMD.
      • by nemequ ( 888009 ) on Tuesday January 02, 2018 @06:15PM (#55851989)
        What are you talking about? That patch says AMD CPU's aren't vulnerable:

        AMD processors are not subject to the types of attacks that the kernel page table isolation feature protects against. The AMD microarchitecture does not allow memory references, including speculative references, that access higher privileged data when running in a lesser privileged mode when that access would result in a page fault.

        The (trivial) patch essential disables the work-around on AMD CPUs. I'm not going to comment on how fair GP's criticism of Intel is in general, but that link definitely isn't evidence in Intel's favor

        • by scumdamn ( 82357 ) on Tuesday January 02, 2018 @10:16PM (#55853173)
          I'm saying that Intel tried to cripple all processors but the dude from AMD told him their products weren't vulnerable and didn't need to be slowed down.
          • by nnull ( 1148259 ) on Wednesday January 03, 2018 @03:35AM (#55854127)
            This, I don't completely understand the reasoning for crippling all processors, including non-intel. It seems Intel is trying to use its political clout to reduce everyone's performance with a bunch of fear and scare tactics which just tops the charts for 2017 and 2018, just so they don't lose their edge. This is just an utter catastrophe for Intel and they're trying to drag the rest of us with them.
            • by zifn4b ( 1040588 )
              Hanlon's Razor: Never attribute to malice that which can be adequately explained by stupidity
      • by sjames ( 1099 ) on Tuesday January 02, 2018 @06:33PM (#55852087) Homepage Journal

        That bolster's AC's point. It looks like the Intel guys were going to cripple performance for everyone until the patch from AMD removed the unnecessary crippling from AMD processors.

  • by jader3rd ( 2222716 ) on Tuesday January 02, 2018 @05:46PM (#55851761)
    Sorry for the lack of imagination, but if the user space process can only read kernel memory, and can't write to it, how could one make use of this?
    • by OverlordQ ( 264228 ) on Tuesday January 02, 2018 @05:48PM (#55851777) Journal

      You're running in EC2 on shared hardware. Your instance can read the memory of other instances running on the same physical hardware.

      • Re: (Score:2, Interesting)

        by Anonymous Coward

        You're running in EC2 on shared hardware. Your instance can read the memory of other instances running on the same physical hardware.

        Yep, "the cloud" bites again.

        When are you supposedly "technically sophisticated" people going to learn that security inherently means running on your own hardware?

        When you cheap out, you lose. Performance, security, integrity, reliability - everything.

        VM specifically: For most local-system applications, VM is a long in the tooth, unneeded hangover that drops performance. I hav

    • by pereric ( 528017 ) on Tuesday January 02, 2018 @06:00PM (#55851893) Homepage

      Cryptographic keys, information on other processes (making other attacks feasible), perhaps random number generator seeds and status, for example ...
      And the principle in general that there could be information the process is not supposed to reach.

    • by EndlessNameless ( 673105 ) on Tuesday January 02, 2018 @06:09PM (#55851955)

      Private keys for system-level crypto and user credentials are stored in kernel space. You want everyone on the system to be reading those? If you can read a private key or a Kerberos token, you can become that daemon/system/user.

      This bug essentially destroys local security and severely compromises network security, subject to any limitations on where/when data can be read.

      I'm not a microarchitecture guru who can dig through the details and figure out the limitations of potential attacks. Perhaps only a small portion of kernel memory can be exposed via this bug. I don't really know. The naive, simple scenario where all kernel memory is exposed, though---that is pretty damned bad. Infosec doomsday bad.

  • by rwven ( 663186 ) on Tuesday January 02, 2018 @05:46PM (#55851763)

    I find it hard to believe that a virtual memory change will result in a 5-30% slowdown for Intel processors. Maybe for a few extremely specific (likely edge-case) tasks, but if there was a legitimate 5-30% performance decrease, you can bet there would be a far different solution in the works that would suitably fix the problem.

    • By "virtual memory" are we talking Page Files and swap space? Disk space as memory?

      So an almost unusuable computer becomes completely unusable. Unless you're on solid state, then you get the performance of a mechanical hdd.
      • by darkain ( 749283 )

        Page File is only one area that can be mapped into a Virtual Address Space. System RAM is another. Often time, I/O is as well.

        https://en.wikipedia.org/wiki/... [wikipedia.org]

      • Re: (Score:2, Interesting)

        by Anonymous Coward

        By "virtual memory" are we talking Page Files and swap space? Disk space as memory?

        No. On modern processor running in protected mode, all memory is virtual. Memory pages that appear contiguous could be backed up by NON-contiguous areas of physical memory, device i/o space, numa, swap, etc. This bug may affect any memory access in userspace.

        So an almost unusuable computer becomes completely unusable.

        OK about the second part. The first part is a matter of opinion ;-)

      • by MrKaos ( 858439 ) on Tuesday January 02, 2018 @11:46PM (#55853505) Journal

        By "virtual memory" are we talking Page Files and swap space?

        Almost. The difference is a minor or major page fault. From swap space to ram to CPU cache is a major page fault whereas a memory transfer is between the CPU cache (L1,L2 or L3) and system RAM is a minor page fault.

        Disk space as memory?

        No. Between CPU cache and system RAM. In this case the issue relates to minor page faults which occur when the CPU scheduler is switching tasks, called "context switching", between processes, threads and lwt running on the system.

        The kernel maintains a summary page of the process when it switches tasks so that it doesn't have to recreate details of the process and is more efficient when it switches tasks. It is this summary page that can be attacked and IIUC addresses can be changed to access memory the process does not have permission to access.

        This means that a process running in the userspace of Ring 3 of the CPU it can modify it's summary table to access data in Ring 0, where the kernel is. Which is obviously bad because now that process can access all the memory.

        So an almost unusuable computer becomes completely unusable. Unless you're on solid state, then you get the performance of a mechanical hdd.

        Maybe, but not how you would expect. The CPU scheduler *may* make it possible to hide some of the latency created by the now crippled context switching in mechanical disk latency that you can't do with ssd because of the way IO determines *when* the scheduler will context switch. Still early days and it depends on what techniques can be devised. It depends on where the summary tables are maintained and I would expect that to be in CPU cache, which is much faster than system RAM, which is why it is a tough bug to get around.

        With that in mind it's plain to see why the Linux kernel devs want to call the patch fuckwit_ because Intel screwed up badly on this.

    • by jonwil ( 467024 )

      If the choice is a 30% slowdown or a massive highly dangerous security flaw, the developers will pick the 30% slowdown. Especially if that flaw is a big problem for people using VMs (there are suggestions in some places that the flaw would be a HUGE problem for cloud providers like Amazon). That said, if you are running Linux and dont care about the security flaw but do care about the slowdown, you can always compile your own kernel without the relavent bits in there :)

      • I can see it being a problem for shared providers etc. HOWEVER, I know a lot of cases in the customer I am working with they will happily accept the risk rather than take an even 5% performance hit as they use the machines for dedicated tasks so if rogue code was somehow on their they would have already been seriously compromised. So I seriously hope regardless of OS that their is a way to disable the fix easily while staying in a supported state if the performance hits are more than just FUD.
        • by sjames ( 1099 )

          This is a good point. If the machine lives in a protected environment where only approved software is used by authorized users, disabling the fix to avoid the slowdown might be the right thing.

          But I'm pretty sure the slowdown in this case isn't FUD. Otherwise we'd hear Intel loudly denying it by now.

      • by Zocalo ( 252965 ) on Tuesday January 02, 2018 @06:20PM (#55852009) Homepage
        Linux users don't even need to compile a custom kernel (although if performance *really* matters you should probably be doing so as a matter of course) as there's a boot time option in the that can be set to disable the new Page Table Isolation mode, "nopti". Without knowing the performance hit for a specific usage case and the nature of the flaw its currently impossible to say whether using it is going to be a good idea or not, but it's nice that they at least thought to include the option. Pretty sure BSD will do the same, but feel free to place your bets on commercial operating system vendors...
      • Re: (Score:2, Insightful)

        If the choice is a 30% slowdown or a massive highly dangerous security flaw, the developers will pick the 30% slowdown.

        As it is it seems a lot of developers choose "30% slowdown" over "spend some time to write not shit code".

        "Premature optimization is the root of all evil" gets turned to "Do no optimization whatsoever, have no understanding of underlying hardware, and pick the latest trendiest framework that runs on top of 5 layers of framework to provide 6502 performance from an i7."

        • by dryeo ( 100693 )

          provide 6502 performance from an i7

          Good, it's a step in the direction of improving latency. Next, replace the rest of the computer with an Apple //E and we'll have something responsive.
          Article measuring latency of various computers that finds the Apple //E to have the least latency in displaying a character, https://danluu.com/input-lag/ [danluu.com]

    • by Carewolf ( 581105 ) on Tuesday January 02, 2018 @05:56PM (#55851843) Homepage

      I find it hard to believe that a virtual memory change will result in a 5-30% slowdown for Intel processors. Maybe for a few extremely specific (likely edge-case) tasks, but if there was a legitimate 5-30% performance decrease, you can bet there would be a far different solution in the works that would suitably fix the problem.

      Virtual memory access is used in every single memory access cached or not. 5% would be lucky for trying to work around a broken system. I am guessing the flaw is probably in the TLB which is meant to accelerate these things.

    • Re: (Score:2, Offtopic)

      by epine ( 68316 )

      I find it hard to believe that a virtual memory change will result in a 5-30% slowdown for Intel processors. Maybe for a few extremely specific (likely edge-case) tasks, but if there was a legitimate 5-30% performance decrease, you can bet there would be a far different solution in the works that would suitably fix the problem.

      Of course, if microcode update fails, there's always the hail Mary unicorn ass-pull.

      I assure you, every Intel employee is kneeling on the carpet this very instant, facing the most aus

    • you cut 30% off the performance of my CPU expect to hear about it.
      • by Xenx ( 2211586 )
        Technically speaking, they're making a change to the operating system kernel and how it operates, not reducing performance of the CPU.
      • by Pyramid ( 57001 )

        "you cut 30% off the performance of my CPU expect to hear about it.
        --
        Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-... [mozilla.org]"

        I can't be the only person who sees the irony of a person complaining about performance degradation and that they make Firefox plug-ins in the same post.

    • Mail everybody a new chip from the past decade would be the "different" solution!.

    • by paskie ( 539112 )

      The slowdown is on syscalls. So it depends on your workload. For example, `du -s` is reportedly slowed down really by tens of percents.

    • by sjames ( 1099 ) on Tuesday January 02, 2018 @07:11PM (#55852257) Homepage Journal

      They don't have a choice. The cost is quite believable since the workaround involves mapping the kernel in and out of the process space for every system call. Keeping it mapped in and keeping the page tables hot in the cache helps performance a lot.

      The real fix involves new silicon.

      • Note that this wouldn't be the first time that Linux has done this. RedHat kernels used to support a 4GB/4GB address split on i386 PAE systems. Conventionally, on 32-bit systems, the kernel reserves the top 2GB of the address space and userspace gets the bottom 2GB. When this started to get cramped, the was an option to reserve the top 1GB for the kernel and make the bottom 3GB available for userspace. Sometimes, even that wasn't enough, so with a 4GB/4GB address split both get completely separate page
    • by grimr ( 88927 ) on Tuesday January 02, 2018 @07:58PM (#55852523)

      I don't think you understand how drastic this fix is. Every time a user mode to kernel mode transition happens and every time a hardware interrupt happens, the entire page table directory layout has to be switched. This means all the TLB caches are flushed as well and that's where the main performance hit comes from.

      So if you're doing something like crypto currency mining you're not going to see much of a hit. But if you're doing a lot of I/O (file servers, database servers, web servers, etc.) you're going to see that 25-35% performance hit.

      And that's why hardware bugs are so serious. Sometimes you get lucky and it's a microcode update with no penalty. Sometimes it's a simple fix with barely any performance penalty. But sometimes you get unlucky and the fix hurts a lot and the only way to get the performance back is to swap out the hardware.

    • Re: (Score:3, Informative)

      by Anonymous Coward

      Every process runs in virtual address space--EVERY process. Low level device drivers on X86 might need to map to physical, but nearly everything is virtual. Certainly everything in user-mode.

      This is why process 1 memory location 10000 is a different piece of physical ram from process 5 memory location 10000. Different virtual address spaces. Each such mapping between each virtual address space and physical space is stored in a "page table". (Pages might also not currently exist in physical ram and the

  • by Artem Tashkinov ( 764309 ) on Tuesday January 02, 2018 @05:47PM (#55851769) Homepage

    The developers behind the GRSecurity project measured up to 63% performance loss [twitter.com]. If most common tasks are equally affected, Intel is sure fucked. Home users might not need to bother, but large cloud providers might be seriously affected.

    Meanwhile the Linux kernel has received the largest incremental minor patch [kernel.org] in its history (229KiB) - perhaps kernel 4.14.11 already contains all the required fixes.

    I have a sneaking suspicion Intel shares will fall through the floor in the next few weeks because Intel CPUs might have suddenly become quite slower than their AMD Zen based counterparts.

    • by Artem Tashkinov ( 764309 ) on Tuesday January 02, 2018 @05:52PM (#55851805) Homepage
      Some PostgreSQL results have just been released [postgresql.org]: up to 23% performance loss. This is indeed huge.
    • or heck if you've just got a low end laptop?
      • I'm already getting serious slowdown from games optimizing for other Windows operating systems. Rebooting into Windows 7 offered an incredible leap in performance on the same rig for one game in particular, Makes the game as fast as it was when I first installed it on the old OS! That is what I get for playing an MMORPG.
    • by Motherfucking Shit ( 636021 ) on Tuesday January 02, 2018 @10:40PM (#55853273) Journal

      I have a sneaking suspicion Intel shares will fall through the floor

      Intel's CEO agrees; a couple weeks ago he sold all the Intel stock he can [fool.com]. If he'd dumped any more shares, he would have had to forfeit his job. That isn't a man who's confident about the future of his company...

      • by Zocalo ( 252965 ) on Wednesday January 03, 2018 @05:18AM (#55854355) Homepage
        This flaw has been known about since at least October, and probably even sooner since early code to fix it starting coming out in November, which makes it seem *highly* likely that he was aware of the it and the potential impact when he took the decision to commit to the November sale. Also, as CEO, there's no way that he can plausibly claim "I didn't know about it" like the two Equifax execs who executed a stock sale just before their breach announcement did without coming across as completely incompetent and unfit for the CEO role. This reeks of insider trading, and after the outcry over the Equifax execs I can't imagine that the SEC isn't going to want to take a good hard look at this stock sale as well.

        I'm sure his vested options will pay for some really good lawyers but, even so, forfeiting his job could be the least of his problems.
  • by sl3xd ( 111641 ) on Tuesday January 02, 2018 @05:52PM (#55851801) Journal

    Linux Weekly News [lwn.net] has been covering this for quite a while.

    5% slowdown on average, with up to 30% for some particularly bad network operations.

    ARM64 is also affected [lwn.net], so it's not just intel

  • by NuclearCat ( 899738 ) on Tuesday January 02, 2018 @05:57PM (#55851859) Journal
    And what is interesting, AMD is immune to that, proof: https://lkml.org/lkml/2017/12/... [lkml.org]
  • AMD is safe (Score:5, Informative)

    by TeknoHog ( 164938 ) on Tuesday January 02, 2018 @06:00PM (#55851899) Homepage Journal

    The summary is not fully explicit: this is not a flaw in Intel x86 ISA, but specific to CPUs made by Intel. AMD processors don't have the problem, so they should not need the patch.

    https://lkml.org/lkml/2017/12/... [lkml.org]

    This could be a huge win for AMD, because the patch incurs a measurable slowdown. At the moment, though, the Linux fix doesn't seem to distinguish between manufacturers. I expect the distinction will appear later -- better safe than sorry.

    • Although AMD is safe, why do they mention a 50% slowdown for AMD processors?

      > @grsecurity measured a simple case where Linux “du -s” suffered a 50% slowdown on a recent AMD CPU.

      • by ColaMan ( 37550 )

        Presumably they enabled the software workaround and ran it on a bunch of CPUs. Then they picked the most alarming slowdown they could find, regardless of whether that CPU needed the workaround or not.

        Or perhaps they were just unaware at the time that AMD CPUs are not at risk.

  • by trybywrench ( 584843 ) on Tuesday January 02, 2018 @06:02PM (#55851911)
    some of my sys admin friends posted this on a slack channel i'm in, apparently it's a big deal

    http://pythonsweetness.tumblr.com/post/169166980422/the-mysterious-case-of-the-linux-page-table [tumblr.com]
  • Anyone remember the TLB bug that also resulted in huge performance penalties in the first generation Phenoms? Guess it's Intel's turn.
  • The future (Score:5, Interesting)

    by Artem Tashkinov ( 764309 ) on Tuesday January 02, 2018 @06:44PM (#55852127) Homepage

    I'm curious how much Cannon Lake and Ice Lake CPU architectures are going to be delayed. Since Cannon Lake is basically SkyLake on a 10nm node, Intel cannot release it with such a glaring hole which causes such a significant performance loss.

    I've been running a Sandy Bridge CPU for seven years now, and now I'm really looking forward to the second gen Zen CPUs. Viva, competition. I'm really glad AMD is still around.

    • by sl3xd ( 111641 )

      I'm curious how much Cannon Lake and Ice Lake CPU architectures are going to be delayed.

      I'm going to go out on a limb and say "not at all"

      CPU design pipelines are pretty long; generally requiring at least a year to go from "Tape Out [wikipedia.org]" to fabrication.

      Releasing no chip (and staying with an even slower current generation) is just not an option.

      Cannon Lake and Ice lake are still an improvement on Intel's current offering, and can still compete against AMD's offering.

      Intel managed to move on after the (even more dire) disasterous NetBurst architecture [wikipedia.org]; there's no reason to believe they won't get pas

  • by Nos. ( 179609 ) <andrew@thekerrsBALDWIN.ca minus author> on Tuesday January 02, 2018 @06:56PM (#55852181) Homepage

    https://www.fool.com/investing... [fool.com]

    Less than a month before we know the linux kernel was being patched for this bug.

    • by sl3xd ( 111641 ) on Tuesday January 02, 2018 @07:27PM (#55852347) Journal

      This bug has been known and reported about since early November; the original paper was presented in July of 2017, and code has been in Github since Feburary.

      Motley Fool is just noting that the Intel CEO isn't holding any more stock than he needs to.

      And there are good reasons:

      * AMD is back from the dead.
      * Intel's GPU hasn't been that successful -- they've even teamed up with AMD to put Radeon GPU's in the same die as an Intel CPU.
      * PC sales are declining as consumers shift from Intel PC's to using ARM-powered tablets & phones instead.
      * ARM is making inroads into the "desktop and laptop computer" marketplace.
      * ARM is powering most consumer electronics as well (TV's, Blu-ray players, Smart Speakers, etc)
      * Intel is absolutely nowhere in the mobile world. Mobile has one ARM to rule them all.
      * Intel missed the boat for the current generation of XBOX and PlayStation consoles.

      Intel is looking more and more like a one trick pony, and its competitors are beginning to do that one trick better too.

  • by nbvb ( 32836 ) on Tuesday January 02, 2018 @07:08PM (#55852247) Journal

    This is why we run our mission critical workloads on SPARC and Power along side Linux. Solaris and AIX. Diversity -- in operating system, in processor, in manufacturer - is healthy. The SPARC T8's are blazing faster, secure, and don't have this nonsense. Neither do our POWER8's. Having all your eggs in the Intel+Linux basket could be a major shitshow here... meanwhile, we'll keep chugging along.

  • Have to ask, does this also affect Intel-Macs? I infer "yes", but have not read many of the detailed articles yet...
    • With 99.99% certainty unless the Mach kernel does something no other OS does in order to secure kernel memory pages.
    • Cue the lawyers to initiate class action lawsuits against Apple once they release their patches to deliberately slow down older machines in the face of a hardware limitation.

  • Go ARM Go (Score:2, Interesting)

    by smist08 ( 1059006 )
    Glad my latest computer is a Raspberry Pi. Glad to be on an ARM processor. Perhaps this will help more ARM based computers become more mainstream this year.
  • The notion that Intel even has the capability of producing new fixed CPUs to match other than the latest packaging/pin requirements seems fanciful. In which case we'll just have to live with any slowdown. As buying all new systems is just too expensive.

  • by bill_mcgonigle ( 4333 ) * on Tuesday January 02, 2018 @08:42PM (#55852713) Homepage Journal

    From the AMD commit [lkml.org]:

    AMD processors are not subject to the types of attacks that the kernel
    page table isolation feature protects against. The AMD microarchitecture
    does not allow memory references, including speculative references, that
    access higher privileged data when running in a lesser privileged mode
    when that access would result in a page fault.

    this can probably be rewritten in the inverse like:

    Intel processors ... allow memory references, including speculative references, that
    access higher privileged data when running in a lesser privileged mode, [including]
    when that access would result in a page fault.

    So it seems like: set up a speculative memory reference to a kernel memory structure, cause a page fault, and then get a bit of kernel memory out (and back in?). That could get you root before long. Some people have been saying this can be leveraged to get a guest into its hypervisor too.

If it wasn't for Newton, we wouldn't have to eat bruised apples.

Working...