Announcement

Collapse
No announcement yet.

The Disappointing Direction Of Linux Performance From 4.16 To 5.4 Kernels

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

  • #31
    Michael

    Any chance you could repeat the same benchmark on the R9 3900X system?

    Comment


    • #32
      Originally posted by Mario Junior View Post

      I'm not a programmer (I don't know even how to write "hello world" on any programming language) and I can compile a kernel.
      So...?
      So can I...well, depending on the package management and how the OS likes to build stuff. Suse f'n confuses me and I don't like that build server crap.

      I can only write "hello world" in shell script, Python, & Rust.

      Granted, any moron should be able to figure out print("hello world"), echo "hello world", & cargo new myProject --bin .

      Comment


      • #33
        Originally posted by Azrael5 View Post
        So what the fuck is happened from the 4.16 to 5.4 kernel?
        Kernel 5.4 isn't even finalized yet. A good bisect would narrow it down. It's very possible Michale is hitting something CPU or workload-specific.


        Linux is doing all is possible to be the worst among the operating system platforms.
        Could you please stop with the "Linux sucks" trolling? Better yet, stop typing or go away if you don't have anything useful to add. You're just making noise and shitting up threads.

        Comment


        • #34
          Typo:

          Originally posted by phoronix View Post
          Like the code comilation slowdown,

          Comment


          • #35
            Originally posted by jrch2k8
            this are hardware issues that will allow attackers to go insanely deep inside your whole system not you run of the mill crappy pointer leaking
            Originally posted by jrch2k8 View Post
            Also depending on your firewall hardware(specially ARM based ones) those vector are exposed as well which with specially crafted packets could exploit other systems as well, also depending on the hardware some of those attack can even compromise the CPU at microcode level or worse.
            Can you point me towards a source describing a Spectre-like attach that "compromises CPUs at microcode level"? Because I've got no idea what you're talking about.

            INB4 JS VM exploit that hops off to the kernel using a local privilege escalation vulnerability, then updates your motherboard's BIOS or CPU's microcode with an unsigned image. Or maybe stealing your Git SSH key through Spectre for someone to put malware in your code without you realizing.
            Last edited by GrayShade; 11-11-2019, 02:13 PM.

            Comment


            • #36
              Originally posted by jrch2k8 View Post

              Well everyone with some mental sanity prefers security over speed when the security breach is that massive, remember this are hardware issues that will allow attackers to go insanely deep inside your whole system not you run of the mill crappy pointer leaking and a brutally dumb patch afterwards.

              If you don't give a rat ass about security on your system just pass mitigations=off at boot and you are golden but don't expect everyone to follow that short sighted idea of yours of let half the planet earth exposed to something this massive because your benchmarks look ugly.
              Security is relative, it can be low or high, but it will never be perfect, no matter how many measures and even if the performance is 1% compared to before.
              Since my computer is not keeping the secrets of NASA or bank deposits, I don't want my performance to be reduced more 10% for protecting nothing so important.
              How about a slider where the user can decide the comporomise between security and performance?

              Plus, when it comes to the security of Linux desktop, In my opinion it's either close to 0 (zero) or just a joke.
              Which Linux distro offers an option to control the installed programs, where I can say what they are allowed to do and what they are not allowed to do?
              Which Linux distro offers and options to control the privacy part of security of the installed programs where I can say from which devices (webcam, mike, scanner, hard disks, flash drives) they are allowed to read from and from which they are not allowed?
              Which Linux distro offers a easy to use firewall where I can say which programs are allowed to use the network and which are not?
              Even Windows has a firewall popup asking to allow or deny programs wanting to access the network.

              So I care about security, but I wish the most important things would be solved first like the control / sandboxing of the installed programs.
              I'm tired of praying that the installed programs will behave nice when I run them or insert the requested superuser password and they get god-like powers and they can destroy everything if they want.
              Why Android can have this and Linux can't.
              Where is the so called security of Linux?

              So I'm sorry, but I'm tired of these security measures on on top of another that just reduce the performance of Linux desktop.
              How about investing more into the other type of security that I said above?

              Comment


              • #37
                I have to admit, I'd like to see 5.4 run with "mitigations=off" as well, as a baseline.

                Comment


                • #38
                  Isn't 4.18 when CoC was inserted into the Linux kernel?

                  Comment


                  • #39
                    Originally posted by Mario Junior View Post

                    I'm not a programmer (I don't know even how to write "hello world" on any programming language) and I can compile a kernel.
                    So...?
                    I can program a PM octree-based 3D navigation database, but I can't compile a kernel and would never want to. :C

                    Comment


                    • #40
                      Headline should read "Current LTS release (4.19) out-performs all other kernels from 4.16 to unreleased 5.4. As expected".

                      Are you trolling the insecure newbies today Michael? Lot of folks peeing their pants today in the comments.

                      Comment

                      Working...
                      X