Announcement

Collapse
No announcement yet.

The Linux 3.0 Kernel Will... Reboot Better?

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

  • The Linux 3.0 Kernel Will... Reboot Better?

    Phoronix: The Linux 3.0 Kernel Will... Reboot Better?

    Beyond file-system cleancache support, a Microsoft Kinect driver, Intel Ivy Bridge support, and various open-source graphics driver improvements, the Linux 3.0 kernel may also reboot your system better. Yes, really...

    http://www.phoronix.com/vr.php?view=OTUwOA

  • #2
    Linux needs to be rebooted?

    News to me.

    Comment


    • #3
      Originally posted by LLStarks View Post
      Linux needs to be rebooted?

      News to me.
      Your aging, security-hole-ridden kernel and low-level user-space would probably appreciate it if you did a yum/apt-get upgrade and rebooted.

      It's like those fools who brag about how their "not IIS" webserver hasn't been restarted in two years. 15 seconds at apache.org, php.net, etc. gives you a laundry list of remote holes to gleefully exploit, if you're the mischievous type.

      Comment


      • #4
        Ksplice does all of that work without forcing a reboot.

        Comment


        • #5
          The article, throughout being rather long and consiting mostly of quotes, doesn't cover the contents of mjg59's blog post to the slightest degree. Michael please try to understand the matter you are posting about, before writing the article.

          Comment


          • #6
            Originally posted by Lynxeye View Post
            The article, throughout being rather long and consiting mostly of quotes, doesn't cover the contents of mjg59's blog post to the slightest degree.
            What? The article was fine.

            Comment


            • #7
              I sometimes wish Michael would just syndicate these blog posts into his site rather than trying to rewrite them. Articles on phoronix should be thought out researched and infromative, not a link fest with the words "leveraged" and "etc." added.

              Comment


              • #8
                it turns out that on most systems the ACPI reboot vector points at 0xcf9 in system IO space.
                first: what is a reboot vector? I know what reboot means, I know what vector means...but throwing those words together doesn't make any sense to me.

                second why is it bad that it points @ 0xcf9? would be 0x1337 more appropriate?

                and what's the big deal about poking the keyboard controller? If I want my PC to reboot I want it to reboot. It should not poke itself o_0

                That's like saying your going to sleep and then, just before sleeping you masturbate. oh wait...

                Comment


                • #9
                  Originally posted by RealNC View Post
                  What? The article was fine.
                  So what does i.e.
                  Originally posted by Article
                  And, shockingly, it turns out that on most systems the ACPI reboot vector points at 0xcf9 in system IO space.
                  say to you?

                  Without the context given by the original blog post this are just words clung together. Random quotes from blogposts you dont understand don'tmake for a good article.

                  Comment


                  • #10
                    So... the default behavior is now to assume that you have broken hardware and treat it as such. Why not just feed the kernel a parameter that says "my hardware is a POS" to enable such moronic behavior?

                    Comment


                    • #11
                      good to see the boot/reboot behavior finally getting some love in the linux desktop.

                      with this and what systemd brings am becoming much happy.

                      Comment


                      • #12
                        Originally posted by droidhacker View Post
                        So... the default behavior is now to assume that you have broken hardware and treat it as such. Why not just feed the kernel a parameter that says "my hardware is a POS" to enable such moronic behavior?
                        Sadly, broken hardware (or broken BIOS that causes broken-hardware-like behavior) is the safe bet when it comes to PC. Perhaps not for any specific piece of hardware, but for a randomly-chosen PC there's a good chance that it has at least one quirk calling for a similarly ugly workaround. If the workarounds were all configurable, configuring a custom kernel would go from "slightly tedious" to "black art".

                        Comment


                        • #13
                          Originally posted by Detructor View Post
                          first: what is a reboot vector? I know what reboot means, I know what vector means...but throwing those words together doesn't make any sense to me.

                          second why is it bad that it points @ 0xcf9? would be 0x1337 more appropriate?

                          and what's the big deal about poking the keyboard controller? If I want my PC to reboot I want it to reboot. It should not poke itself o_0

                          That's like saying your going to sleep and then, just before sleeping you masturbate. oh wait...
                          If you read the blog post, you'll find out why they point to 0x0CF9. It's bad because it's an undocumented feature, not standardised, and strictly speaking requires a pair of magic words, though appears to work with only one written twice. That's how I read it anyhow.
                          Basically the way to signal a reset/reboot to the system is about as murky and abhorrent as the wireless mbus standard.

                          Comment

                          Working...
                          X