Announcement

Collapse
No announcement yet.

Deleting A Few Lines Of Code Can Yield "Significant" Power Savings On Some Linux Systems

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

  • #31
    Originally posted by lowlands View Post

    Not much if not insignificant when put in perspective: how much power could have been saved by *not* mining bitcoin?
    but apparently bit coint had a purpose compare to wasting power for nothing

    Comment


    • #32
      Originally posted by ix900 View Post

      If they fully tested everything and fully checked everything, it would be too slow, and production would be really low. They would need a lot more people, a lot more testers, etc. Not in this world.
      they should do just that considering that the kernel is the thing that manage your hardware. you can fix a software error but a hardware crash is not that easy

      Comment


      • #33
        Originally posted by loganj View Post

        but apparently bit coint had a purpose compare to wasting power for nothing
        Blockchain certainly has its merits but IMHO Bitcoin serves no purpose and only wastes gigantic amounts of energy.

        Comment


        • #34
          Originally posted by nuetzel View Post
          Still compiling (amd-staging-drm-next (5.6.0)) with these lines deleted.
          Probably you don't need to, try to pass to the kernel: pcie_aspm=force

          Comment


          • #35
            Hopefully I see some saving from this patch on my i5-3437U.
            Code:
            lspci | grep bridge
            00:00.0 Host bridge: Intel Corporation 3rd Gen Core processor DRAM Controller (rev 09)
            00:1c.0 PCI bridge: Intel Corporation 7 Series/C216 Chipset Family PCI Express Root Port 1 (rev c4)
            00:1c.2 PCI bridge: Intel Corporation 7 Series/C210 Series Chipset Family PCI Express Root Port 3 (rev c4)
            00:1c.3 PCI bridge: Intel Corporation 7 Series/C216 Chipset Family PCI Express Root Port 4 (rev c4)
            00:1f.0 ISA bridge: Intel Corporation QM77 Express Chipset LPC Controller (rev 04)

            Comment


            • #36
              Originally posted by kokoko3k View Post
              Probably you don't need to, try to pass to the kernel: pcie_aspm=force
              But that's my normal process during amdgpu bug hunting...

              Comment


              • #37
                Aside from Phoronix, do any of you know of any other site, Twitter account, YouTube channel or any other on-line publication that regularly performs detailed and thorough benchmarks of the Linux kernel (as well as many other prominent open source software) and carefully publishes the resulting stats in detail?

                I'm not aware of any.

                Linux is the most important OS kernel in the word, yet the only one I know of who regularly tests and benchmarks it in public is Michael Larabel. More than once, he was the first to notice and report major performance regressions in newer kernels, with the developers fixing them thanks to his hard work.

                And all these years he's being doing that thankless work all by himself, using his home's basement as a test lab, underfunded and without going on vacation for years. And he's been maintaining his benchmarking software as an open source project for everyone to use as well.

                Michael, your work often goes underappreciated, but the Linux kernel would be in a worse shape than it is today if it wasn't for you.

                Here's a shoutout to you, brother. 🤗

                You should consider starting a YouTube channel with a Patreon account, though. You might reach a wider audience that way and perhaps earn more revenue as well.

                Comment


                • #38
                  Originally posted by Dark-Show View Post
                  Hopefully I see some saving from this patch on my i5-3437U.
                  Code:
                  lspci | grep bridge
                  00:00.0 Host bridge: Intel Corporation 3rd Gen Core processor DRAM Controller (rev 09)
                  00:1c.0 PCI bridge: Intel Corporation 7 Series/C216 Chipset Family PCI Express Root Port 1 (rev c4)
                  00:1c.2 PCI bridge: Intel Corporation 7 Series/C210 Series Chipset Family PCI Express Root Port 3 (rev c4)
                  00:1c.3 PCI bridge: Intel Corporation 7 Series/C216 Chipset Family PCI Express Root Port 4 (rev c4)
                  00:1f.0 ISA bridge: Intel Corporation QM77 Express Chipset LPC Controller (rev 04)
                  It seems you have NO PCIe - PCI/PCI-X bridges.

                  Comment


                  • #39
                    Originally posted by nuetzel View Post

                    It seems you have NO PCIe - PCI/PCI-X bridges.
                    Well that's a disappointment, thanks for the clearification, I was purely looking for PCI bridges.

                    Comment


                    • #40
                      Originally posted by andrei_me View Post

                      Wasn't Apple that had a security bug because of an "if" without braces?
                      Possibly, that might have been an NSA plant.

                      In any event the interesting thing here is that Linux never really strived to be a low power solution, whereas Apple pulled out all the stops possible to lower system power usage. So now we have a UNIX like solution that urns on a cell phone fairly well. Linux development is more focused on the server space thus less of an interest in things that save power. Given all of that I would not be surprised if Apple found a long standing issue causing excessive power usage in their kernels.

                      I have to laugh at the shock some are expressing here. There will always be long standing bugs that when fixed solve a problem people didn't even knew existed.

                      Comment

                      Working...
                      X