Announcement

Collapse
No announcement yet.

Linux Kernel Set To Finally Retire AMD 3DNow!

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

  • #41
    Originally posted by Anux View Post
    Whoever wants to limit the waste he produces will use its tools till they can't be repaired. Buying something new and throwing away what still works causes more energy consumption and pollution.
    unless that new is much less power hungry(which it will be if new enough)

    Comment


    • #42
      Originally posted by Anux View Post

      That might be true if money is your only concern. Whoever wants to limit the waste he produces will use its tools till they can't be repaired. Buying something new and throwing away what still works causes more energy consumption and pollution.

      My home NAS has 30 - 35W idle (old Phenom2 quad core) but it can deliver > 100 MB/s filetransfers with full disk encription on raid 5. I have a Raspi3 that i tested to replace it, but its NAS capabilities are weak ~ 10MB/s without raid and encription. Thats to slow if you really need a NAS, not to speak of multiple people using it at the same time.
      But here my Phenom is as slow as a Raspberry Pi 4B... Yep, 50MB/s over SFTP...

      Comment


      • #43
        Originally posted by willmore View Post
        So we add features with hundreds of KLOC, but we remove a small chunk of well tested code because...progress?
        Evidently the code wasn't that well tested since it was broken. From the commit message:

        This code puts an exception table entry on the PREFETCH instruction to overwrite it with a JMP.d8 when it triggers an exception. Except of course, our code is no longer writable, also SMP.

        Instead of fixing this broken mess, simply take it out.

        Comment


        • #44
          Originally posted by pal666 View Post
          unless that new is much less power hungry(which it will be if new enough)
          You totally missed my point, you not only need to account for energy consumption while using hardware but also the energy that was used to produce an ship it, as well as the disposal/recycling process. https://en.wikipedia.org/wiki/Agbogbloshie

          Originally posted by tildearrow View Post

          But here my Phenom is as slow as a Raspberry Pi 4B... Yep, 50MB/s over SFTP...
          Than your storage device seems to be the limit and your still better off not producing more waste if it doesn't improve your experience.

          Comment


          • #45
            Originally posted by rmfx View Post

            Yep, there are millions of other bs lines of code to remove. Like the graphic drivers. A light and universal cpu vulkan support a la lavapipe should be enough to drive graphics until mesa or blobs are installed.
            I am not an expert on the graphics stack but this does not make sense to me. The only parts of those drivers that reside inside the kernel are the lightweight parts..... Besides, why turn all the drivers into userspace drivers? This would drop gpu performance and (re-)create other issues.

            Comment


            • #46
              Originally posted by ResponseWriter View Post

              You could argue that about a lot of things. I don't, for example, use or need AVX512. Like I said, the hardware supports SSE4a so for me the loss of 3DNow! support will hardly be noticed.
              AVX512 is far more powerful and modern, supported in modern hardware, and likely to be introduced to almost all cpu architectures in the near future. AMD will include it.... And Alder Lake shenanigans aside, Intel is likely to re-introduce it in consumer cpus as well. 3DNow is totally useless in modern use cases and used only in obsolete AMD processors. Apples to oranges comparison.

              Comment


              • #47
                Originally posted by Nuc!eoN View Post
                With the global chip shortage, and no end in sight it is beyond me why support/features for older hw generations is getting removed from the linux kernel and mesa recently.
                Yeah. People stuck on 2013-15 eras hardware are going to miss 3Dnow for their use cases....

                Comment


                • #48
                  Originally posted by Anux View Post
                  Than your storage device seems to be the limit and your still better off not producing more waste if it doesn't improve your experience.
                  Nope, as transfer using netcat uses the available throughput entirely (which is 110 MB/s) unlike SFTP. Here:

                  Code:
                  sftp> put mmc-4.gz
                  Uploading mmc-4.gz to /location/mmc-4.gz
                  mmc-4.gz                                                                      100%  611MB  65.9MB/s   00:09     
                  sftp> ^D
                  ~$ cat mmc-4.gz | pv | nc ??? 4567  
                  611MiB 0:00:05 [ 108MiB/s] [         <=>                                                                      ]
                  Last edited by tildearrow; 14 December 2021, 04:44 AM.

                  Comment


                  • #49
                    Originally posted by tildearrow View Post

                    Nope, as transfer using netcat uses the available throughput entirely (which is 110 MB/s) unlike SFTP. Here:

                    Code:
                    sftp> put mmc-4.gz
                    Uploading mmc-4.gz to /location/mmc-4.gz
                    mmc-4.gz 100% 611MB 65.9MB/s 00:09
                    sftp> ^D
                    ~$ cat mmc-4.gz | pv | nc ??? 4567
                    611MiB 0:00:05 [ 108MiB/s] [ <=> ]
                    You know about caching? You're a member of a forum all about benchmarks after all and wouldn't make such a big mistake right?
                    Last edited by Anux; 14 December 2021, 06:33 AM.

                    Comment


                    • #50
                      Originally posted by Nuc!eoN View Post
                      With the global chip shortage, and no end in sight it is beyond me why support/features for older hw generations is getting removed from the linux kernel and mesa recently.
                      In this case, if you look at the commit the article links to, the old 3dnow code had bitrotten and was broken. By removing that code and falling back to the default x86-32 memory copying and page clearing functions (that ALL other x86-32 CPU's use, so they are very likely correct and decently well performing), the commit actually made the kernel work again for those old CPU's.
                      Last edited by jabl; 14 December 2021, 06:27 AM.

                      Comment

                      Working...
                      X