Announcement

Collapse
No announcement yet.

Intel's Mitigation For CVE-2019-14615 Graphics Vulnerability Obliterates Gen7 iGPU Performance

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Intel's Mitigation For CVE-2019-14615 Graphics Vulnerability Obliterates Gen7 iGPU Performance

    Phoronix: Intel's Mitigation For CVE-2019-14615 Graphics Vulnerability Obliterates Gen7 iGPU Performance

    Yesterday we noted that the Linux kernel picked up a patch mitigating an Intel Gen9 graphics vulnerability. It didn't sound too bad at first but then seeing Ivy Bridge Gen7 and Haswell Gen7.5 graphics are also affected raised eyebrows especially with that requiring a much larger mitigation. Now in testing the performance impact, the current mitigation patches completely wreck the performance of Ivybridge/Haswell graphics performance.

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Said on the last post about this exploit but might as well pop it here as well.
    I wonder if this patch can be worked to disable on the fly and have Feral's GameMode / Steam gain a massive new Intel feature?
    Might be better than forgetting to re-enable the patch and being compromised from browsing, if you're a gamer.

    On second thought, the disable process could probably then be exploited. Let's hope for an alternative (faster) patch.
    Last edited by Dark-Show; 15 January 2020, 09:28 PM.

    Comment


    • #3
      Originally posted by Dark-Show View Post
      Said on the last post about this exploit but might as well pop it here as well.
      I wonder if this patch can be worked to disable on the fly and have Feral's GameMode gain a massive new Intel feature?
      Might be better than forgetting to re-enable the patch and being compromised from browsing, if you're a gamer.
      As said in the article, there currently is no toggle for disabling the patch (short of not running Gen7 graphics). And as said, hopefully there will be such an option though for the kernel module parameter. It's unlikely though to be a on-the-fly parameter like would be handled by GameMode but likely boot-time only.
      Michael Larabel
      https://www.michaellarabel.com/

      Comment


      • #4
        Good grief, no thanks.

        Comment


        • #5

          Comment


          • #6
            Originally posted by Michael View Post

            As said in the article, there currently is no toggle for disabling the patch (short of not running Gen7 graphics). And as said, hopefully there will be such an option though for the kernel module parameter. It's unlikely though to be a on-the-fly parameter like would be handled by GameMode but likely boot-time only.
            Or you could buy an AMD based system. Hell you could buy an ARM based laptop and get better results. Maybe ARM is a stretch due to there not being any laptops worth buying yet but I think my point is clear.

            Your testing indicates more damage than I first suspected, it makes me wonder how visible this will be in the corporate world where the sales are right now for Intel.

            Comment


            • #7
              FYI, will also be testing on some other Ivy bridge / Haswell systems tomorrow, gotta see what I still have in the racks (or what I feel like assembling otherwise when digging through the CPU vault).
              Michael Larabel
              https://www.michaellarabel.com/

              Comment


              • #8
                Ugh by 2050 our processor will be as slow as a 68000 due to mitigation overload :<

                Comment


                • #9
                  Originally posted by tildearrow View Post
                  Ugh by 2050 our processor will be as slow as a 68000 due to mitigation overload :<
                  Moore's inverse law is now coming into effect do to the non-flat nature of our cosmos. Processing power will now continuously contract do to the nature of dark-mitigation until a singularity is reached and all CPUs implode. As this is a geometric acceleration process I think the world will end as we know it no later than 2025.

                  Comment


                  • #10
                    Originally posted by tildearrow View Post
                    Ugh by 2050 our processor will be as slow as a 68000 due to mitigation overload :<
                    If you have a motherboard still capable of running the remaining couple of Ivy Bridge era CPUs - due to tin whisker death - you may not be able to tell the difference in any software on the 68k and the Ivy Bridge... think about how little current (and historical) software developers care about efficient resource use.

                    Comment

                    Working...
                    X