Announcement

Collapse
No announcement yet.

Linus Torvalds Is Doing More ARM64 Linux Testing Now That He Has A More Powerful System

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

  • DanaG
    replied
    I hope he'll try a Navi GPU on the thing and see the amdgpu driver fail to load firmware.
    On my ARM64 machine (LX2160A), it seems like the driver is assuming that all architectures behave the same as x86 around alignment and PCIe write-combining.
    Last edited by DanaG; 21 May 2024, 05:15 PM.

    Leave a comment:


  • ayumu
    replied
    Originally posted by mrg666 View Post

    Reading this from Qualcomm, I think I agree with you.
    https://www.qualcomm.com/news/onq/20...-its-potential
    And now that RISE has been around for a while, we can feel its effects. They already have a history of valuable contributions.



    Leave a comment:


  • mrg666
    replied
    Originally posted by ayumu View Post

    Implies there have been such announcements and they weren't true? I have no such memory.

    RISC-V is inevitable. Think about it carefully, one company's roadmap might slip up. All of them? Doubtful.
    Reading this from Qualcomm, I think I agree with you.

    Leave a comment:


  • varikonniemi
    replied
    Originally posted by sophisticles View Post

    Do you know what "airgapping" is? It's the computer not being connected to an outside network.

    The compromises that you are referring to where most likely an inside job, money talks and it speaks louder than ideology.
    airgapping is compatible with usb sticks being used which in that case caused the infection. Isolating means nothing touches your computer that has not been explicitly forensically declared safe.

    Leave a comment:


  • Jonjolt
    replied
    Originally posted by sophisticles View Post

    Do you know what "airgapping" is? It's the computer not being connected to an outside network.

    The compromises that you are referring to where most likely an inside job, money talks and it speaks louder than ideology.
    I heard the DoD even requires the electrical system to be set up to prevent side channel attacks.

    Leave a comment:


  • Chugworth
    replied
    Originally posted by sophisticles View Post

    You do realize that the nature of his work, creating GPL'd software, means that he has no secrets, everything he does is literally out in the open, hence the term "open source".

    Why bother tricking him into accepting compromised hardware when you can just go to git and see all the code?
    The concern is not simply the open source code itself. State-level actors could try to intercept and manipulate private data and communications or attempt various means of blackmail.

    Leave a comment:


  • sophisticles
    replied
    Originally posted by varikonniemi View Post

    nice theory. But in practice that level of paranoia helps nothing if your attacker also controls the router and bypasses your fancy whitelist.

    Even airgapping helps little against a state level attacker, as we saw with iranian nuclear centrifuges selfdestructing. Only staying isolated from the world will keep you safe.
    Do you know what "airgapping" is? It's the computer not being connected to an outside network.

    The compromises that you are referring to where most likely an inside job, money talks and it speaks louder than ideology.

    Leave a comment:


  • sophisticles
    replied
    Originally posted by Chugworth View Post
    It's anyone's guess how well his computer security is. But when a person has a high level of influence, hacking attempts can be much more sophisticated and targeted. It would definitely be a good idea to be cautious with any equipment that's received for free.
    You do realize that the nature of his work, creating GPL'd software, means that he has no secrets, everything he does is literally out in the open, hence the term "open source".

    Why bother tricking him into accepting compromised hardware when you can just go to git and see all the code?

    Leave a comment:


  • Weasel
    replied
    RISC-V is a joke, the other benchmark clearly shows this and yet there's still shills who cope so hard to expect it to be good, as you can see from this thread.

    Pure cope.

    Leave a comment:


  • torsionbar28
    replied
    Now how long until there is a powerful RISC-V system that attracts Torvalds' interest?
    How many powerful desktop RISC-V systems are available? Unless Linus drives a certain Delorean at 88 miles per hour, hardware that doesn't exist yet, is rather difficult to obtain.

    Leave a comment:

Working...
X