Announcement

Collapse
No announcement yet.

Linux KVM Virtualization Had Mistakenly Been Applying L1TF Workaround To Unaffected CPUs

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

  • #11
    Originally posted by ddriver View Post

    IIRC that very explicit clarification was not part of the initial article, must have been added in light of my comment. At any rate, I cannot help but feel curious how exactly memory encryption related code ended up triggering a redundant security issue mitigation. That meticulous "intel didn't do it" is not suspicious in the least...

    Also F much? That's bad karma right there
    My position is that words are just words and that everyone has a trigger so it really doesn't matter what words are used to describe something since someone will take offense.

    Wanna see some Phoronix-style bad karma?

    Comment


    • #12
      Originally posted by cynic View Post

      what's wrong with his name?
      There's nothing wrong with his name, it's a truly cool name, and he's a cool guy for finding this stuff and fixing it.

      Comment


      • #13
        QEMU 5.0 broke host-passthrough, and it doesn't look like it's being addressed any time soon. So now I have to run my four Windows 10 VMs in crippled host-model mode. I know QEMU is not KVM, but supposedly the issue has something to do with the way QEMU 5.0 interfaces with KVM.

        Does anyone know the details about this problem, or the ETA for a fix?

        Comment


        • #14
          Originally posted by gukin View Post

          There's nothing wrong with his name, it's a truly cool name, and he's a cool guy for finding this stuff and fixing it.
          oh, sorry! I thought you were sarcastic
          Indeed, Paolo is great!

          Comment


          • #15
            I read the article. I see that AMD messed up their own stuff. Mistakes happen...

            ...but I thought the Linux development process included code reviews by interested parties and developers/maintainers that are responsible for that branch. I saw that level of review activity years ago in the LM Sensors code branch, when I was very interested in that sort of stuff.

            At least, post mortem, a developer/maintainer that oversees KVM code for Linux upstream (per the article) and also works for Redhat fixed the borkage.

            So I just have to ask...where was this Redhat person when this code was originally submitted by KVM? Asleep on a beach some place without Internet access? or they suddenly had a flash of insight and learning regarding L1TF code and impacted processors?

            And why is that question important?

            I honestly see a complete breakdown of the code review process that should be going on here with KVM code that is upstreamed to Linux mainline.

            To me it's both funny and sad when I see Linux developers/maintainers with high levels of responsibility for upstream Linux code do stuff that resembles tying their shoelaces together.

            Comment


            • #16
              Originally posted by skeevy420 View Post
              ...
              Your mom called from wccftech, they need you back. She said she knows you're bottom-hurt from intel's engineering performance of late, but they will probably eventually be fine, in a few years, maybe, so you should go back home where you truly belong.

              Now seriously - it is true that some people have been having all sorts of ridiculous and ungrounded problems with words and general freedom of expression. This is a sinister agenda to limit the dynamic range of human expression, and has practically nothing to do with your already flattened to an angstrom expressability. Adding in redundant curse words that add no information to language, and to excess at that, in a way that isn't even funny or remotely original, all because you are unintelligent enough to actually believe it to be cool. Acting out is such a juvenile manner clearly proves that if anyone here is being triggered, that is you. That being said, I respect your right to be a joke, but I am not sure you respect, or for that matter even acknowledge your right not to be one. Food for thought. Apologies if my sentences are a bit too long for your attention span, but true communication does not end at mindless tropes and memes.


              Originally posted by NotMine999 View Post
              I read the article. I see that AMD messed up their own stuff. Mistakes happen...
              You see that from where? From the article stating it for a fact or from researching the matter? Because I am still waiting for the factual substantiation of that weird "it wasn't intel, it was amd" part.

              Comment


              • #17
                Originally posted by ddriver View Post
                ...
                Are you aware that some people consider direct attacks and bringing someone's mom into it as much worse than using the word fuck?




                Comment

                Working...
                X